Commodity Prices and Development

Commodity Prices and Development This page intentionally left blank Commodity Prices and Development Edited by Roma...

Author: Roman Grynberg | Samantha Newton

115 downloads 2140 Views 2MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Commodity Prices and Development

This page intentionally left blank

Commodity Prices and Development Edited by Roman Grynberg and Samantha Newton

1

3

Great Clarendon Street, Oxford ox2 6dp Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With ofﬁces in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries Published in the United States by Oxford University Press Inc., New York ß Commonwealth Secretariat 2007 The moral rights of the author have been asserted Database right Oxford University Press (maker) First published 2007 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this book in any other binding or cover and you must impose the same condition on any acquirer British Library Cataloguing in Publication Data Data available Library of Congress Cataloging in Publication Data Data available Typeset by SPI Publisher Services, Pondicherry, India Printed in Great Britain on acid-free paper by Biddles Ltd., King’s Lynn, Norfolk ISBN 978–0–19–923470–7 1 3 5 7 9 10 8 6 4 2

Contents

List of Figures List of Tables Notes on Contributors Introduction Roman Grynberg and Samantha Newton

vii xi xiii 1

PART I The Issue of Declining Commodity Prices 1. The Problems of Commodity Dependence Mohammad A. Razzaque, Philip Osafa-Kwaako, and Roman Grynberg

7

2. Secular Decline in Relative Commodity Prices: A Brief Review of the Literature Mohammad A. Razzaque, Philip Osafa-Kwaako, and Roman Grynberg

17

3. Long-Run Trend in the Relative Price: Empirical Estimation for Individual Commodities Mohammad A. Razzaque, Philip Osafa-Kwaako, and Roman Grynberg

35

4. Analysis of Movements in the Productivity and Prices of Selected Tropical Commodities in Developing Countries, 1970 to 2002 Euan Fleming, Prasada Rao, and Pauline Fleming 5. Commodity Value Chains Compression—Coffee, Cocoa, and Sugar Jaya Choraria

PART II

68

136

The Implications of Declining Commodity Prices

6. Estimating Foreign Exchange Loss due to Declining Commodity Prices Mohammad A. Razzaque, Philip Osafa-Kwaako, and Roman Grynberg

163

v

Contents 7. Marginalization of LDCs and Small Vulnerable States in World Trade Bijit Bora, Roman Grynberg, and Mohammad A. Razzaque

PART III

175

Mitigating the Impacts for Commodity Dependent Countries

8. Instruments for Addressing Commodity Price Behaviour Mohammad A. Razzaque, Philip Osafa-Kwaako, and Roman Grynberg

269

9. Commodity Prices and the Debt Relief Initiative Mohammad A. Razzaque, Philip Osafa-Kwaako, and Roman Grynberg

278

10. Aid Flows and Commodity Prices Mohammad A. Razzaque, Philip Osafa-Kwaako, and Roman Grynberg

301

References Index

329 345

vi

List of Figures

1.1 Share of Agricultural Products in Global Merchandise Exports

10

1.2 Per Capita Exports and Primary Exports as Percentage of Merchandise Exports in 144 Developing Countries

10

1.3 Share of LDCs, SVs, and HIPCs in World Merchandise Exports: 1950–2001

12

1.4 Relationship between Real GDP Growth Rate and Share of Primary Exports in Total Merchandise Export Volume in Developing Countries

12

2.1 Grilli–Yang Relative Price of Primary Commodities and its Changes Over Time

24

3.1 Relative Prices of 13 Commodities: 1900–2001

41

3.2 Real Prices of Broad Commodity Groups

46

3.3 Estimated Growth in Relative Prices for Broad Commodity Groups

46

3.4 Trend (1960–2002) Growth Rates for Individual Commodities with UNCTAD Data

49

3.5 Trend Growth Rates since the 1960s: Grilli–Yang versus UNCTAD Data

50

4.1 Export Quantity Index for Selected Commodities in All Sampled Countries, 1970–2002

75

4.2 Export Quantity Index for Selected Commodities in Sampled Commonwealth Countries, 1970–2002

76

4.3 Export Quantity Index for Selected Commodities in Sampled African Countries, 1970–2002

76

4.4 Export Quantity Index for Selected Commodities in Sampled African Commonwealth Countries, 1970–2002

77

4.5 Production Functions, Technological Change, and Technical Efﬁciency Change

81

4.6 Export Price Index for Selected Commodities in All Sampled Countries, 1970–2002

96

4.7 Export Price Index for Selected Commodities in Sampled Commonwealth Countries, 1970–2002

97

4.8 Export Price Index for Selected Commodities in Sampled African Countries, 1970–2002

97

vii

List of Figures 4.9 Export Price Index for Selected Commodities in Sampled African Commonwealth Countries, 1970–2002

98

4.10 Export Price and Import Price Indices for All Commodities in All Sampled Countries, 1970–2002

99

4.11 Export Price and Import Price Indices for Tree Crops in All Sampled Countries, 1970–2002

99

4.12 Export Price and Import Price Indices for Field Crops in All Sampled Countries, 1970–2002

100

4.13 Annual Rates of Change in Labour Productivity in All Sampled Countries, 1970–2002

102

4.14 Annual Rates of Change in TFP in All Sampled Countries, 1970–2002

103

4.15 Annual Rates of Change in TFP in Commonwealth Countries, 1970–2002

106

4.16 Trends in Export Unit Values and TFP in Jamaica, 1970 to 2002

116

4.17 Trends in Export Unit Values and TFP in Fiji, 1970 to 2002

117

4.18 Trends in Export Unit Values and TFP in Solomon Islands, 1970 to 2002

117

4.19 Trends in Export Unit Values and TFP in Ghana, 1970 to 2002

118

4.20 Selected Countries with Lower Rate of TFP Growth to Rate of Decline in Export Unit Value, 1970 to 2002

120

4.21 Trends in Export Unit Values and TFP in Malaysia, 1970 to 2002

121

4.22 Selected Countries Experiencing Rates of Decline in Export Unit Values and TFP, 1970 to 2002

121

4.23 Trends in the Single Factoral Terms of Trade in Nigeria, 1970 to 2002

124

4.24 Trends in the Single Factoral Terms of Trade in the Central African Republic, 1970 to 1998

125

4.25 Trends in the Single Factoral Terms of Trade in Papua New Guinea, 1970 to 1998

126

4.26 Trends in the Single Factoral Terms of Trade in Costa Rica, 1970 to 2002

127

4.27 Trends in the Single Factoral Terms of Trade in Mauritius, 1970 to 1998

127

4.28 Trends in the Single Factoral Terms of Trade in Solomon Islands, 1970 to 1998

128

4.29 Trends in the Single Factoral Terms of Trade in Sri Lanka, 1970 to 2001

128

4.30 Trends in the Single Factoral Terms of Trade in Kenya, 1970 to 1998

129

4.31 Trends in the Single Factoral Terms of Trade in Sierra Leone, 1970 to 1998

130

4.32 Trends in the Single Factoral Terms of Trade in Trinidad and Tobago, 1970 to 1998

130

5.1 Coffee: Cameroon-UK

151

5.2 Coffee: Ethiopia-UK

151

viii

List of Figures 5.3 Coffee: Kenya-UK

152

5.4 Coffee: PNG-UK

152

5.5 Coffee: Tanzania-UK

153

5.6 Coffee: Ghana-UK

153

5.7 Sugar: farm gate-to-retail price spreads

154

5.8 Sugar: farm gate-to-retail price spreads

154

5.9 Sugar: Mauritius-US

155

5.10 Sugar: Mauritius-EU

155

5.11 Sugar: Fiji-US

156

5.12 Sugar: Fiji-EU

156

5.13 Sugar: Brazil-US

157

5.14 Sugar: Thailand-US

157

5.15 Sugar: Australia-US

158

6.1 Composite Relative Commodity Price Index and its Changes

166

6.2 Volume and Purchasing Power of Exports

168

6.3 Foreign Exchange Loss as a Percentage of Primary and Merchandise Exports

171

7.1 Share of LDC Exports in Global Merchandise Exports, 1950–2000

182

7.2 Share of Small States in Global Merchandise Exports, 1950–2000

182

7.3 Share of Small States and LDCs in Commercial Services Exports

186

7.4 Declining Importance of Small States and LDCs in World Export (Merchandise Plus Services) Trade

187

7.5 Share of Small States and LDCs in World Trade Transactions

188

7.6 Aggregate Exports (Merchandise Plus Services) of Individual LDCs ($million)

190

7.7 Aggregate Exports (Merchandise Plus Commercial Services) of Individual Small States

191

7.8 Share of Individual LDCs in World Aggregate Exports, 1980–2000

192

7.9 Share of Individual Small States in Aggregate Global Exports, 1980–2000

193

7.10 Marginalization of Individual LDCs in Aggregate Exports (Merchandise plus Commercial Services), 1980–2000

197

7.11 Marginalization of Individual Small States in Total Exports (Merchandise plus commercial services), 1980–2000

197

7.12 Net Shifts in 1995–2000 as Percentage of 1990–94 Average Exports (Merchandise Plus Services) for LDCs

206

7.13 Net Shifts in 1995–2000 as Percentage of 1990–94 Average Exports (Merchandise Plus Services) for Small States

206

7.14 Trends in Marginalization and Growth of Real GDP in LDCs

207

ix

List of Figures 7.15 Trends in Marginalization and Growth of Real GDP in Small States

207

7.16 Composition of Exports in LDCs: Primary vs. Manufacturing

209

7.17 Share of Primary and Manufacturing Exports in Small States

209

7.18 Share of Agriculture in World Exports and World Exports-GDP Ratio

212

7.19 Scatter Plot of lnMAR and lnAGX

213

7.20 Scatter Plot of lnMAR and lnGLO

214

7.21 Plot of Variables and their Correlograms

217

7.22 Scatter Plot of lnMARSS and lnAGX for Small States

223

7.23 Scatter Plot of lnMARSS and lnGLO for Small States

223

7.24 Share of LDCs and Small States in Global Inﬂow of FDI

228

9.1 Real Commodity Price Index and Real Outstanding Debt in HIPCs

281

9.2 Projected Export Growth and NPV Debt-to-Export Ratio for Countries that have Reached Completion Point

289

9.3 Average 1990–99 Actual Export Growth Rate vis-a`-vis Projected Growth Rate for HIPCs

291

10.1 Composite Relative Commodity Prices and Aid Flows to LDCs, 1980–2000

303

10.2 Composite Relative Commodity Prices and Aid Flows to HIPCs, 1980–2000

303

10.3 Composite Relative Commodity Prices and Aid Flows to Small States, 1980–2000

304

10.4 Aid Flows to Mali and Cotton Prices

304

10.5 Aid Flows to Papua New Guinea and the Real Cocoa Price

305

10.6 Aid Flows to Togo and the Real Phosphate Price

305

x

List of Tables

1.1 Commodity Export Dependence in LDCs, SVs, and HIPCs

8

1.2 Large Share of Export Earnings from a Single Commodity in LDCs, SVs, and HIPCs

9

2.1 Summary of Findings on Secular Decline in Commodity Prices

31

3.1 Regression Results for 13 Commodities (with Updated Grilli-Yang Series: 1900–2001)

44

3.2 Regression Results for Broad Commodity Groups as in UNCTAD Commodity Price Bulletin: Annual Data (1960–2002)

45

4.1 Contributions by Selected Commodities to Export Earnings and Agricultural Output

73

4.2 Estimates of Trends in Export Quantities of Selected Commodities, 1970 to 2002

75

4.3 Major Exports of Selected Commodities

90

4.4 Proportion of the Total Value of Crop Output Contributed by the Selected Commodities in 1990

92

4.5 Estimates of Trends in Export Unit Values of Selected Commodities, 1970 to 2002

95

4.6 Estimated TFP Model

110

4.7 Estimated Labour Productivity Model

112

4.8 Aggregate Rates of Change in TFP and Export Prices

114

4.9 Comparison of Rates of Change in TFP and Selected Commodity Prices

116

4.10 Trends in the Single Factoral Terms of Trade

123

5.1 Breakdown of 2003 Raw Sugar Sales

148

6.1 Estimated Foreign Exchange Loss by Individual LDCs, SVs, and HIPCs (US$ million in 1984–86 prices)

169

6.2 Cumulative Foreign Exchange Loss from some Selected Commodities, 1985–2000

172

7.1 Absolute Volume of Exports

177

7.2 Absolute Volume of Merchandise Imports

179

7.3 Absolute Growth of Merchandise Exports ($billion)

180

7.4 Trend Growth Rates of Exports (per cent)

181

xi

List of Tables 7.5 Exports of Commercial Services ($billion)

183

7.6 Imports of Commercial Services ($billion)

184

7.7 Absolute Growth of Commercial Services Exports ($billion)

185

7.8 Growth Rates of Exports of Commercial Services

185

7.9 Volume of Export Trade (Merchandise Plus Commercial Services) ($ billion)

186

7.10 Total Trade Transactions of Different Country Groups ($billion)

188

7.11 Growth Rates of Merchandise and Services Exports from Individual LDCs

194

7.12 Growth Rates of Merchandise and Services Exports From Individual Small States

195

7.13 Average Change in Exports of LDCs in the 1990s

199

7.14 Average Change in Exports of Small States in the 1990s

200

7.15 A Summary of Trends in Marginalization of LDCs in the 1990s

201

7.16 A Summary of Trends in Marginalization of Small States in the 1990s

203

7.17 Fall in Commodity Prices in Real Terms

210

7.18 Computed F Test Statistics and Critical Values

215

7.19 DF and ADF Tests for Unit Roots

216

7.20 PHFMOLS Estimates of the Model

220

7.21 Short-Run Error Correction Model

222

7.22 Unit Root Test for lnMARSS

224

7.23 Short-Run Error Correction Model

225

7.24 Ofﬁcial Financial Flows ($ million)

227

8.1 Salient Features of Five Important International Commodity Agreements

271

9.1 Foreign Exchange Losses from Commodities and Outstanding Debt

281

9.2 Debt Relief Initiatives

283

9.3 Debt Relief for HIPC Countries

284

9.4 Debt Indicators in Developing Countries and HIPCs, 1999 (%)

286

9.5 Actual and Projected Debt Service Indicators for HIPCs that have Reached Decision Point

294

9.6 Hypothetical Cost of Compensation for HIPCs

297

10.1 Cost Estimates for a Joint Diversiﬁcation Fund for LDCs, HIPCs and Small States (US$ million in 1984–86 prices)

312

10.2 Hypothetical Burden Sharing among Donors

314

10.3 ODA/GNI Positions of Donors after the Hypothetical Contributions to the Joint Diversiﬁcation Fund

315

xii

Notes on Contributors

Bijit Bora

Counsellor, Economic Research and Statistics Division, World Trade Organization, Geneva, Switzerland.

Jaya Choraria

Research Fellow, Economic Affairs Division, Commonwealth Secretariat, London, United Kingdom.

Euan Fleming

Senior Lecturer, Department of Agricultural and Resources Economics, The University of New England, Armidale, Australia.

Pauline Fleming

Lecturer, School of Economics, The University of New England, Armidale, Australia.

Roman Grynberg

Deputy Director, Economic Affairs Division, Commonwealth Secretariat, London, United Kingdom.

Philip Osafa-Kwaako

Research Fellow, Economic Affairs Division, Commonwealth Secretariat, London, United Kingdom.

Prasada Rao

Professor, School of Economics, University of Queensland, Australia.

Mohammad A. Razzaque

Lecturer, Department of Economics, University of Dhaka, Dhaka, Bangladesh.

xiii

This page intentionally left blank

Introduction Roman Grynberg and Samantha Newton

More than ﬁfty developing countries depend on three or fewer commodities for more than half of their exports. Thirty-four of the Less Developed Countries (LDCs) rely on primary commodities to contribute at least half of their export earnings; for seventeen of them, primary commodities contribute more than 75 per cent. Twenty-two Small Vulnerable States (SVs) rely on commodities for more than 50 per cent of exports. Similarly, 32 of the 42 HIPCs are predominantly exporters of primary commodities. In fact, reliance on only a single commodity for a large share of export earnings is quite common in these countries, exposing them to the risk of export earnings instability as a result of price shocks and, perhaps even more signiﬁcantly, falling purchasing power of exports over the long run in the face of the declining real price of the commodity in question. Over the past two decades the prices of nearly all the major agricultural commodities declined in real terms. Whether the terms of trade have moved unfavourably against primary commodities has been a subject of great controversy in the development economics literature since 1950 when Prebisch and Singer ﬁrst hypothesized the problem. Despite contrasting evidence, in recent times there is a broad consensus for the long-run deterioration in relative commodity prices. The research carried out by such inﬂuential multilateral organizations as the World Bank and IMF has contributed to the formation of this consensus position. The studies documented in this book, the outcome of ﬁve years of research at the Commonwealth Secretariat, add further weight to Prebisch and Singer’s hypothesis. The empirical investigation presented in Chapter 3 of this text provides evidence of the presence of a statistically signiﬁcant declining trend in the relative price of most individual commodities, much higher rates of decline being observed for the more recent period. It was found that over the past century, the estimated trend growth rates for most commodities fell between 0.79 and 1.43 per cent per annum. Much higher rates of decline are observed

1

Introduction over the relatively more recent period. Between 1960 and 2002 the aggregate relative price of commodity has fallen at an annual rate of 1.82 per cent with the corresponding ﬁgures for individual commodities ranging from 0.9 to 3.50 per cent. UNCTAD (2004a, p. 22) observed that the net effect of the secular decline in prices depends on two things—the extent to which world market prices are transmitted to producers and whether higher export volumes (eg through productivity and yield improvements) make up for falling prices. Chapter 4 addresses UNCTAD’s later point. The study, focusing on tropical commodity (coffee, cocoa, copra, palm kernel oil, coconut oil, palm oil, rice, cotton and sugar) dependent developing countries, investigates whether producers of commodities in developing countries have compensated for falling producer prices by increasing total factor productivity and whether falling export prices have been compensated for by rising total factor productivity of commodities at the national level in developing countries. It was found that very few of the countries studied had achieved rates of productivity growth that even matched, let alone counteracted, the rate of change in real prices. In determining the extent to which world market prices are transmitted to producers, the issue of a long run secular decline in the relative price of primary commodities must be considered in the context of the issue of a decrease in the producer’s share of retail value over time. Despite numerous quantitative studies providing evidence to illustrate the extent of the problems of commodity prices, historically there has been a lack of quantitative analysis of the evolution of the producer’s share of total retail value. However, Chapter 5 details a study of commodity value chain compression for coffee, cocoa, and sugar. The study uses time series data of prices along entire commodity chains from raw material, in a commodity exporting developing country, to ﬁnal retail product, in a developed consuming country, in order to provide descriptive analysis of the evolution of farm gate-to-retail price spreads. Comparisons are made across the commodities studied and across countries in order to provide insight into the causes of changes in the farm gate-to-retail price spread over time. The evidence gathered on widening farm gate-to-retail price spreads (equivalent to a decrease in the farmer’s share of retail value over time) illustrates the plight of farmers in commodity exporting developing countries. Interestingly, evidence suggests that the compression suffered by sugar farmers in Fiji and Mauritius, countries enjoying preferential trading agreements with the EU, was less severe than for farmers in countries which did not beneﬁt from the Sugar Protocol. The persistent weakness of real commodity prices presents serious challenges for export earnings and domestic incomes in commodity dependent countries. Secular decreases in real prices of commodities have caused lower purchasing power of primary exports, on which most of these countries rely predominantly for ﬁnancing their imports. The resultant foreign exchange

2

Introduction losses relative to the total primary and merchandise exports of many of these countries are quite substantial. It is estimated that during the period 1995–2000, the countries comprising LDCs, HIPCs and SVs, suffered a cumulative foreign exchange loss of US$37 billion due to weakness in commodity prices (i.e. about US$6 billion per annum). For LDCs, the average annual loss is estimated to be about US$2.3 billion while the corresponding ﬁgures for HIPCs and SVs are US$5.5 and US$0.6 billion respectively. For many countries cumulative losses from a single commodity were found to be very large. For commodity dependent poor countries, persistent downward trends in real commodity prices, unallayed by higher export volumes, have resulted not only in signiﬁcant foreign exchange losses but also in a failure to derive much beneﬁt from the ongoing process of trade liberalization and globalization. We have attempted (in Chapter 7) to explain marginalization of LDCs and SVs in merchandise exports in terms of falling share of agricultural products in total global exports and in terms of world export-GDP ratio. The study establishes a valid long-run statistical relationship, indicating that these factors explain about 91 and 85 per cent variation in the declining share of world trade of LDCs and SVS respectively. A review of existing and recent instruments in international commodity policy ﬁnds that these instruments did not address the issue of long-run weakness in primary commodity prices. While price stabilization was the principal motive of the international commodity agreements, nevertheless they attempted, through market intervention, to raise the depressed price levels for a number of commodities. However, since the collapse of commodity agreements, there has not been any signiﬁcant initiative to revive the prices of commodities. IMF external compensatory ﬁnancing and EU-STABEX schemes focused only on the shortfalls in absolute export earnings and export earnings from commodities and commodity prices were not speciﬁcally targeted. On the other hand, various commodity protocols under EU-ACP trade arrangements guaranteed preferential prices for speciﬁc commodities exported by some selected suppliers. The scope of such preferences was very limited; however evidence suggests that farmers of the speciﬁc commodities covered in countries that beneﬁted from the arrangements may have suffered less severe compression of commodity prices than those that did not beneﬁt; e.g. sugar farmers in Fiji and Mauritius. Most commodity dependent low-income countries have also become heavily indebted and are included in the World Bank-IMF sponsored HIPC initiative. While the HICP debt relief initiative is commendable, the failure to address the problems of weakness in commodity prices adversely affecting export earnings prospects of the beneﬁciary countries threatens the credibility of the scheme. It is argued that a permanent solution to the problem of debt crisis lies in the structural shift in composition of the export basket of these countries. We propose an expansion of the HIPC initiative to include all LDCs

3

Introduction and SVCs and a supplementary debt-relief support, which would provide additional debt relief to the HIPCs in the event of adverse trends in commodity prices leading to unsustainable debt burden. Aid ﬂows to LDCs, HIPCs, and SVs have not attempted to compensate for the losses incurred by the recipient countries as a result of the secular decline in commodity prices. Sustained weakness in commodity prices requires export diversiﬁcation and structural changes in the economy. The international community can support the attempts toward diversiﬁcation made by poor, commodity dependent countries. The study thus proposes the establishment of a multilateral fund that would provide resources needed to support diversiﬁcation projects in commodity dependent developing countries. Illustration of hypothetical schemes shows that contribution by the donors to a multilateral diversiﬁcation fund on the basis of some proportion of terms of trade loss suffered by LDCs, HIPCs, and SVs would increase the donors’ current ODA/ GNI ratio only marginally. The problems faced by commodity dependent LDCs, HIPCs, and SVs are formidable. Although diversiﬁcation is the most appropriate response to the problem of the secular decline in commodity prices, long-term transformation in the economy can be a slow process and in the long-run the success will depend on a host of such factors as the development of human resources, institutional capacity building, poverty alleviation, and appropriate domestic policy and environment. By granting increased aid ﬂows and debt relief, and providing assistance to encourage production of non-traditional export items, the international community can play a proactive role in the development of the commodity dependent poor countries. Only concerted efforts both at the domestic fronts of these countries and co-operation extended by the international community can help mitigate the problem of the world’s most vulnerable economies.

4

Part I The Issue of Declining Commodity Prices

This page intentionally left blank

1 The Problems of Commodity Dependence Mohammad A. Razzaque, Philip Osafa-Kwaako, and Roman Grynberg

For a long time, commodity prices have been a source of considerable interest among academic researchers, and have been a major cause of concern for policy-makers and a harsh reality in the lives of poor people in countries that rely predominantly on primary production and exports. Primary commodity prices are not only associated with violent ﬂuctuations, but have also exhibited a long-run declining trend relative to manufactured goods. When the declining terms of trade for primary commodities was ﬁrst catapulted into prominence (Prebisch, 1950 and Singer, 1950), concerns were expressed that it would lead to unequal distribution of gains from trade between the primary producing developing countries and the developed economy suppliers of manufactured goods. In today’s world, however, the exports of developing countries as a group are dominated by manufactured items and consequently commodity production does not act as a divide between the North and the South.1 Nevertheless, for the overwhelming majority of the economies classiﬁed, not mutually exclusively, as least developed countries (LDCs), small vulnerable states (SVs) and heavily indebted poor countries (HIPCs), dependence on the production of primary commodities remains a major challenge for development.2 In 34 LDCs (about 70 per cent), primary commodities contribute at 1 Industrialized countries are still the dominant suppliers of primary commodities in the world market. According to UNCTAD data, in 2000 developed market economies accounted for 64 per cent of world’s primary exports (excluding fuels). Primary exports (without fuels) constituted about 11 per cent of total merchandise exports of both the developed and developing countries. 2 LDCs are considered to be the poorest countries of the world and are frequently deemed to be structurally handicapped in their development. This study uses the December 2001 list of 49 LDCs as deﬁned by the UN Economic and Social Commission. On the other hand, because of the small size of their domestic economies, their remoteness and isolation and their economic vulnerability and susceptibility to natural disaster, SVS are also confronted with

7

The Issue of Declining Commodity Prices Table 1.1. Commodity Export Dependence in LDCs, SVs, and HIPCs Dependence greater than Dependence between 75 % 50–75 %

Dependence between 25–49 %

Dependence less than 25 %

Solomon Islands (96%)*§ Burundi (92%)*y Suriname (92%)§ Uganda (92%)*y Samoa (91%)*§ Ethiopia (90%)*y Niger (90%)*y St Vincent and Grenadines (90%)§ Rwanda (89%)*y Zambia (89%)*y Malawi (88%)*y Tonga (84%)§ Belize (82%)§ Kiribati (82%)*§ Madagascar (81%)*y Nicaragua (81%)y Vanuatu (81%)*§ Congo, D. R. (80%)*y Guyana (80%)§y Jamaica (80%)§ Gambia (79%)*§y Guinea-Bissau (79%)*y Sao Tome and Principe (79%)*§y Guinea (76%)*y

Benin (48%)*y Seychelles (48%)§ Afghanistan (45%)* Cyprus (42%)§ Djibouti (40%)*§ Lao PDR (37%)*y Eritrea (36%)* Haiti (36%)* Mauritius (34%)§ Tuvalu (34%)*§ Barbados (25%)§

Bhutan (19%)* Nepal (19%)* Gabon (18%)§z Bangladesh (16%)* Cambodia (16%)* Botswana (15%)§ Equatorial Guinea (13%)*§ Lesotho (12%)*§ Republic of Congo (8%)y Antigua and Barbuda (7%)§ Trinidad and Tobago (5%)§z Yemen (5%)*zy Malta (4%)§ Angola (2%)*zy

Mauritania (73%)*y Somalia (73%)*y ˆ te D’Ivoire (71%)y Co Ghana (71%)y Tanzania (71%)*y Chad (70%)*y Honduras (69%)y Mozambique (68%)*y Myanmar (68%)*y Papua New Guinea (67%)§ Grenada (66%)§ Maldives (66%)*§ Burkina Faso (65%)*y Kenya (65%)y Togo (65%)*y Bolivia (64%)y Comoros (64%)*§y Mali (64%)*y Central Af. Rep. (63%)*y St Lucia (61%)§ Liberia (60%)*y Sudan (60%)*zy Cameroon (59%)y Swaziland (57%)§ Vietnam (54%)y Dominica (55%)§ Senegal (55%)*y St Kitts and Nevis (53%)§ Fiji (52%)§ Cape Verde (51%)*§ Sierra Leone (50%)*y

Note : *indicates that the country is a least developed country, § a small vulnerable state, z an oil producing country and y a highly indebted poor country. The ﬁgures within the parentheses are average commodity dependence for periods 1980, 1985, 1990, 1995, and 2000. The dependence on primary commodity is estimated excluding the contribution of fuels in total merchandise exports. Source : Authors’ estimates based on data from UNCTAD.

least half of export earnings; for 17 of them, primary commodities contribute more than 75 per cent (Table 1.1). In the case of SVs, there are 22 countries (about 63 per cent of all SVS) where commodities account for more than 50 per cent of exports. Similarly, 32 of the 42 HIPCs (88 per cent) are predominantly overriding problems constraining their economic development. The deﬁnition of a small state covers all countries with a population of less than 1.5 million, and also includes Botswana, Jamaica, Mauritius and Papua New Guinea, even though they have populations above the threshold (Grynberg and Razzaque, 2003). Finally, the group of HIPCs comprises 42 poor countries that have accumulated unsustainable external debt. The deﬁnitions of LDCs, SVs and HIPCs are not mutually exclusive: 13 small states are LDCs of which two are also HIPCs, and 32 LDCs are HIPCs. Only eight HIPCs are neither LDCs nor SVs. Altogether, 81 countries can be considered as either LDCs, small states or HIPCs. Appendix 1.1 gives a list of these countries.

8

Problems of Commodity Dependence Table 1.2. Large Share of Export Earnings from a Single Commodity in LDCs, SVs, and HIPCs Commodities

50 per cent or more

20–49 per cent

10–19 per cent

Crude Petroleum

Angola, Gabon, Republic of Congo, Yemen

Cameroon, Equatorial Guinea, Trinidad and Tobago, Papua New Guinea St Vincent, Honduras Jamaica, Suriname

Vietnam

Bananas Bauxite Cashew Nuts Cocoa Coffee (Arabica) Coffee (Robusta) Copper Copra and coconut oil Cotton Diamond Fish Gold Jute Livestock Iron Ore Rice Sugar

Guinea Guinea Bissau Sao Tome and Principe, ˆ te d’Ivoire, Ghana Co Burundi, Ethiopia

Cameroon Rwanda

Honduras, Nicaragua

Uganda

Cameroon

Zambia

D. R. Congo, Papua New Guinea Solomon Islands

Kiribati

Mauritania

Benin, Chad, Mali, Sudan Central Af. Republic Mozambique Ghana Mali Mauritania Mauritius, Swaziland, Guyana, St Kitts and Nevis

Tea Timber

Tobacco Uranium Vanilla

St Lucia

Equatorial Guinea, Lao PDR, Solomon Islands

Burkina Faso D.R. Congo Senegal, Maldives Mali, Guyana Bangladesh Niger, Sudan, Nicaragua Guyana Belize

Kenya, Rwanda Cambodia, Central Af. Republic, Gabon, Ghana Myanmar, Papua New Guinea, Swaziland

Malawi Niger Comoros

Source : Cashin et al. (1999).

exporters of primary commodities. Not only do these three groups of countries rely heavily on commodities, but their exports are also concentrated either on a single commodity or on a limited range of exports. In 40 countries (out of a total of 81 LDCs, SVs and HIPCs), three leading commodities account for more than 50 per cent of export earnings (Appendix 1.2). Reliance on a single commodity for a large share of export earnings is quite common in these countries (Table 1.2), exposing them to the risk of export earnings instability as a result of price shocks and falling purchasing power of exports over the long run in the face of the declining real price of the commodity in question.

9

The Issue of Declining Commodity Prices 0.2 0.18 0.16

ratio

0.14 0.12 0.1 0.08

1997

1994

1991

1988

1985

1982

1979

1976

1973

1970

0.06

Figure 1.1. Share of Agricultural Products in Global Merchandise Exports Note and source: Agricultural exports data are from FAO Commodity Yearbook (various issues), while the data on merchandise have been taken from UNCTAD (2002).

ln (per capita exports in US$)

12.0

y = −0.021x + 6.4612 R2 = 0.1159

10.0 8.0 6.0 4.0 2.0 0.0 0.0

20.0

40.0

60.0

80.0

100.0

Primary exports as % of merchandise exports Figure 1.2. Per Capita Exports and Primary Exports as Percentage of Merchandise Exports in 144 Developing Countries Note: Oil-rich developing countries have been excluded. Data on per capita exports are for 1998–2000 average. The vertical axis shows the natural logarithm of per capita exports.

10

Problems of Commodity Dependence There are serious problems associated with excessive dependence on commodity production and exports. On the demand side, low-income elasticity of demand for primary commodities, together with technological advances resulting in declining intensity in the use of raw materials, has exerted a downward pressure on the expansion of overall consumption. Indeed, during the past three decades the share of agricultural products in global merchandise exports has more than halved—falling from about 18 per cent in 1970 to less than 8 per cent in 2000 (Figure 1.1). On the supply side, the improvement of technology, the emergence of new suppliers and the agricultural policy of developed countries have contributed to a rapid expansion in world commodity supplies (Reinhart and Wickham, 1994). The resultant imbalance, stemming from the surge in supply vis-a`-vis depressed demand, has caused a secular decline in relative commodity prices. The declining terms of trade would suggest reduced purchasing power of exports of countries predominantly dependent on primary commodities. This problem is further exacerbated by the interaction between price-inelastic and low-income elasticity of demand for commodities. That is, when the demand is not increasing, the revenue from a commodity with price-inelastic demand will fall if supply is increased. The consequence, known as the ‘adding-up’ problem, is that all commoditydependent countries cannot achieve high export growth. The cross-country experience suggests an inverse relationship between the degree of dependence on primary commodities and per capita exports among the set of developing countries (see Figure 1.2). The low price and income elasticity of demand, falling share of agricultural products in global merchandise exports, and ‘adding-up’ problem imply that if a group of countries continues to specialize in primary products, it will be marginalized in world trade. Between 1950 and 2000 LDCs’ share in world merchandise exports thus fell from more than 2.5 per cent to about 0.44 per cent (Figure 1.3). As most HIPC members are also LDCs, the former closely resembles the marginalization trend of the latter. Finally, the share of SVs in world export trade dropped from 0.5 to 0.2 per cent.3 Apart from the declining terms of trade, commodities have experienced widespread shocks in their prices. Typical large negative shocks have been found, with a year-on-year price fall of 44 per cent and a direct loss of income for given export quantities estimated to be 7 per cent of GDP (Collier, 2002).4 On the other hand, positive price shocks are known to have generated 3 It needs to be mentioned here that falling shares in world export volume may not be a problem as long as a country’s exports grow at some fair rate. However, many commoditydependent countries have been subject to frequent falls in absolute export revenues. With already low export volumes, if these countries cannot increase their share in world trade, globalization will only contribute to more skewed distribution of gains from trade. 4 Collier (2002) estimates that each dollar of direct loss from large terms of trade shock costs the economy US$3.

11

The Issue of Declining Commodity Prices 4 3.5 LDCs

per cent

3

SVs

HIPCs

2.5 2 1.5 1

2000

1998

1996

1994

1992

1990

1988

1986

1984

1982

1980

1978

1976

1974

1972

1970

1968

1966

1964

1962

1960

1958

1956

1954

1952

0

1950

0.5

Figure 1.3. Share of LDCs, SVs, and HIPCs in World Merchandise Exports: 1950–2001 Note: Oil-rich countries are excluded. Source: Authors’ estimates.

‘Dutch disease’ effects for non-commodity export and import-competing sectors (Yabuki and Akiyama, 1996). On the whole, the commodity-dependent countries have grown more slowly than others (see Figure 1.4) and the overwhelming majority of them saw declines in purchasing power parity

Real GDP Growth Rate (1980 –2000)

12

y = −0.0275x + 4.6955 R2 = 0.1181

10 8 6 4 2 0 −2 −4 0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

100.0

Primary exports as percentage of Merchandise Exports Figure 1.4. Relationship between Real GDP Growth Rate and Share of Primary Exports in Total Merchandise Export Volume in Developing Countries Note: Based on 116 countries for which data are available.

12

Problems of Commodity Dependence (PPP) adjusted per capita incomes (Birdsall and Hamoudi, 2002).5 In 1999, the average real GDP per capita (adjusted for purchasing power) was lower in nonoil commodity-exporting LDCs than it had been in 1970 (UNCTAD, 2002b). There has also been a clear link between dependence on exports of primary commodities and the incidence of extreme poverty. It has been found that the type of export in which poor countries specialize makes a big difference in their degree of economic success and pattern of poverty. In particular, about Appendix 1.1. List of LDCs, Small States and HIPC Countries Countries

LDCs

Afghanistan Antigua and Barbuda Angola Bahrain Bangladesh Barbados Belize Benin Bhutan Bolivia Botswana Burkina Faso Burundi Cambodia Cameroon Cape Verde Central African Republic Chad Comoros Congo ˆ te d’Ivoire Co Cyprus DR Congo Djibouti Dominica Equatorial Guinea Eritrea Ethiopia Fiji Gabon Gambia Ghana Grenada Guinea Guinea-Bissau Guyana Haiti

Yes

Small States

HIPCs

Yes Yes

Yes Yes

Yes Yes Yes Yes Yes

Yes Yes Yes

Yes Yes Yes

Yes Yes Yes

Yes Yes Yes Yes

Yes

Yes

Yes Yes Yes Yes Yes

Yes Yes Yes Yes Yes Yes

Yes

Yes Yes Yes Yes Yes Yes Yes Yes

Yes Yes

Yes Yes Yes Yes

Yes Yes Yes

Yes (Continued)

5 The lower growth prospect of commodity-dependent economies is often referred to as ‘resource curse’ in development economics literature, where a rich endowment of natural resource is considered to be detrimental to industrialization or even development of institutions. See Bonaglia and Fukasaku (2003) for a review.

13

The Issue of Declining Commodity Prices Appendix 1.1. (Continued ) Countries Honduras Jamaica Kenya Kiribati Lao PDR Lesotho Liberia Madagascar Malawi Maldives Mali Malta Mauritania Mauritius Mozambique Myanmar Nepal Nicaragua Niger Papua New Guinea Rwanda Samoa Sao Tome and Principe Senegal Seychelles Sierra Leone Solomon Islands Somalia St Kitts and Nevis St Lucia St Vincent and the Grenadines Sudan Suriname Swaziland Togo Tonga Trinidad and Tobago Tuvalu Uganda United Republic of Tanzania Vanuatu Vietnam Yemen Zambia

LDCs

Small States

HIPCs Yes

Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

Yes Yes Yes Yes Yes Yes Yes Yes Yes

Yes

Yes Yes

Yes Yes Yes

Yes Yes Yes Yes

Yes Yes Yes Yes Yes Yes

Yes Yes Yes

Yes Yes

Yes Yes Yes Yes

Yes Yes Yes Yes Yes Yes

Yes

Yes Yes Yes

Yes

Yes Yes Yes Yes Yes Yes

Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

Source: The lists of LDCs and HIPCs are from UNCTAD (2002b). Small states include those countries as listed in Grynberg and Razzaque (2003). There are thirteen countries that are both small states and LDCs, viz. Cape Verde, Comoros, Djibouti, Equatorial Guinea, Gambia, Kiribati, Lesotho, Maldives, Samoa, Sao Tome and Principe, Solomon Islands, Tuvalu and Vanuatu. * indicates that the oil-exporting small-LDC, Equatorial Guinea, is not included. Among the 49 LDCs 32, i.e. 65 per cent, are highly indebted poor countries. Four small states viz. Comoros, Gambia, Guyana, and Sao Tome and Principe are also classiﬁed as HIPC countries. Two countries, Gambia and Sao Tome and Principe, have all three characteristics, being LDCs, small states and HIPCs. Bolivia, Cameroon, Congo, Ghana, Honduras, Kenya, Nicaragua, and Vietnam are the only eight countries that do not fall into either LDCs or small states but are HIPCs.

14

Problems of Commodity Dependence Appendix 1.2. LDCs, HIPCs, and Small States and Their Leading Exports Countries

Three Leading Commodities (1997–99)

Afghanistan

Grapes and Raisins, Hides and Skins, Crude Materials (incl. Flowers) Fish, Beverages Dist Alcoholic, Wood Fuels, Diamonds, Coffee Fuels, Iron, Oil Palm Fish, Jute and Bust Fibres, Tea Sugar, Beverages Dist Alcoholic, Fuels Sugar, Bananas, Fish Cotton, Cottonseed, Oil Palm OrangesþTangþClem, Wheat/Flour, Fruit Freshens Oilseed, Fuel, Soybean Oil Diamonds, Bovine Meat, Hides and Skins Cotton, Sesame Seed, Hides and Skins Coffee, Tea, Sugar Wood, Natural Rubber, Fish Fuels, Wood, Cocoa Fish, Apples, Wood Diamonds, Wood, Cotton Cotton, Live Animals, Crude Materials (incl. Flowers) Vanilla, Essential Oils, Cloves (whole þ stems) Fuels, Wood, Sugar Cocoa, Fuels, Coffee Tobacco, Roots and Tubers, Dairy Products Diamonds, Coffee, Wood Sugar, Crude Materials (incl. Flowers), Fish Sugar, Tobacco, Cocoa Fuels, Wood, Cocoa Sesame Seed, Hides and Skins, Fish Coffee, Hides and Skins, Sesame Seed Sugar, Gold, Fish Fuels, Wood, Manganese ore Groundnuts, Fish, Groundnut Oil Cocoa, Diamonds sorted, Gold Nutmeg, Mace, Cardamom, Fish, Wheatþ Flour Bauxite, Alumina (Al Oxide, Hydroxide), Fish Nuts, Fish, Cotton Gold, Sugar, Bauxite Coffee, Fish, Mangoes Coffee, Bananas, Fish Alumina (Al Oxide, Hydroxide), Sugar, Bauxite Tea, Coffee, Fuels Fish, Copra, Crude Materials (incl. Flowers) Wood, Coffee, Tin Ore Wool, Greasy, Food Wastes, Vegetables Prepared Natural Rubber, Wood, Fuels Fish, Coffee, Cloves (wholeþstems) Tobacco, Tea, Sugar Fish, Wood, Copra Cotton, Live Animals, Groundnut Oil

Antigua and Barbuda Angola Bahrain Bangladesh Barbados Belize Benin Bhutan Bolivia Botswana Burkina Faso Burundi Cambodia Cameroon Cape Verde Central African Rep. Chad Comoros Congo ˆ te d’Ivoire Co Cyprus DR Congo Djibouti Dominica Equatorial Guinea Eritrea Ethiopia Fiji Gabon Gambia Ghana Grenada Guinea Guinea-Bissau Guyana Haiti Honduras Jamaica Kenya Kiribati Lao PDR Lesotho Liberia Madagascar Malawi Maldives Mali

Average Share (per cent) (1997–99) 48.58 2.57 71.00 66.36 8.65 19.44 52.48 37.86 7.42 23.31 73.20 41.45 88.91 40.45 44.10 21.54 73.15 52.44 65.48 85.83 60.00 43.14 86.25 7.22 34.11 89.06 6.11 79.42 33.85 93.22 19.03 61.88 23.17 59.92 75.42 90.98 18.68 34.89 61.15 46.07 80.38 10.32 3.02 14.56 54.19 70.96 71.70 45.13 (Continued)

15

The Issue of Declining Commodity Prices Appendix 1.2. (Continued ) Average Share (per cent) (1997–99)

Countries

Three Leading Commodities (1997–99)

Malta Mauritania Mauritius Mozambique Myanmar Nepal

Tobacco, Fish, Beverages Dist Alcoholic Iron ore and concentrates, Fish, Fuels Sugar, Fish, Crude Materials (incl. Flowers) Fish, Nuts, Wood Wood, Fish, Pulses Roots and Tubers, Pulses, Nutmeg, Mace, Cardamom Coffee, Fish, Bovine Meat Uranium, Live Animals, Tobacco Gold, Copper ore, Wood Coffee, Tea, Hides and Skins Fish, Copra, Fruit Prepared Cocoa, Fish, Coffee Fish, Fuels, Groundnut Oil Fish, Fuels, Cinnamon (Canella) Fish, Coffee, Cocoa Wood, Fish, Oil of Palm Live Animals, Bananas, Fish Sugar, Beverages Non-Alcoholic, Beverages Dist Alcoholic Bananas, Fruit Fresh, Pepper (White/Long/ Black) Bananas, WheatþFlour, Rice

1.14 72.43 23.28 42.92 45.08 8.97

Sesame Seed, Crude Materials (incl. Flowers), Coarse Grains Alumina (Al Oxide, Hydroxide), Rice, Fuels Sugar, Fruit Prepared, Other Citrus Fruits Nat. Ca Phosphate, Cotton, Coffee Pumpkins, Squash, Gourds, Fish, Crude Materials (incl. Flowers) Fuels, Beverages Non-Alcoholic, Sugar, Copra Coffee, Fish, Crude Materials (incl. Flowers) Nuts, Coffee, Fish

28.69

Copra, Roots and Tubers, Wood Fuels, Rice, Fishery Commodities Fuels, Fish, Coffee Reﬁned Copper, Sugar, Cotton

67.02 32.76 89.30 49.61

Nicaragua Niger Papua New Guinea Rwanda Samoa Sao Tome and Principe Senegal Seychelles Sierra Leone Solomon Islands Somalia St Kitts and Nevis St Lucia St Vincent and the Grenadines Sudan Suriname Swaziland Togo Tonga Trinidad and Tobago Tuvalu Uganda United Republic of Tanzania Vanuatu Vietnam Yemen Zambia

36.67 93.70 57.92 69.62 76.43 92.11 41.72 35.00 26.61 95.13 41.20 36.61 55.68 68.48

84.50 23.24 61.75 82.02 51.77 16.30 65.94 42.17

Source : UNCTAD database.

four-ﬁfths of extremely poor people live in those least developed countries that are mainly primary producers (UNCTAD, 2002a). The probability of becoming heavily indebted is also higher for commodity-producing poor countries. About 50 developing countries depend on three or fewer commodities for more than half their exports; 37 of these have been categorized as HIPCs.

16

2 Secular Decline in Relative Commodity Prices: A Brief Review of the Literature Mohammad A. Razzaque, Philip Osafa-Kwaako, and Roman Grynberg

Whether the terms of trade have moved unfavourably against primary commodities and the developing countries dependent on them has been the subject of intense interest and debate in the trade and development literature since the publication of articles by Prebisch (1950) and Singer (1950) some 53 years ago. The issue of movement of the terms of trade is essentially an empirical question and the hypothesis of a long-term trend decline in relative commodity prices has been the subject of one of the liveliest debates in the empirical economics literature. Statistical and econometric tests have been applied to produce evidence and counter-evidence. As Sapsford and Balasubramanyam (1994) appositely observe: ‘ . . . declining long-run trend hypothesis has in recent years established itself as an important test bed, upon which time series statisticians nowadays routinely evaluate their latest trend estimation procedure’.1 The basic objective of this chapter is to provide a brief review of the literature concerning the secular decline in commodity prices in an attempt to recapitulate the main issues in the controversial empirical research works and to identify any broad consensus that may have appeared in the recent past.2 A summary of selected studies highlighting the methodologies and data used, as well as key ﬁndings, is presented in Table 2.1.

1

Sapsford and Balasubramanyam (1994), p. 1737. Important previous reviews include Greenaway and Morgan (1999), Sapsford and Balasubramanyam (1994), Sapsford and Morgan (1994) and Sapsford and Singer (1998). Note that we restrict ourselves to the literature on secular decline in relative prices. Other issues, such as volatility and co-movement of prices, are not addressed. 2

17

The Issue of Declining Commodity Prices

2.1. The Genesis of the Debate Classical economists predicted a long-run improving trend in the prices of primary commodities relative to those of manufactures.3 According to the classical view, primary commodity production tends to be subject to diminishing returns and technological progress is likely to be more rapid in manufacturing than in agriculture. If prices are related to costs, the interaction of these two forces will lead the ratio of prices of primary products to those of industrial goods to rise (Thirlwall, 1989). In contrast, Prebisch (1950) and Singer (1950) identiﬁed a number of factors that were considered actually to have contributed to the deterioration of the net barter terms of trade (NBTT) for agricultural products.4 As summarized in Athukorala (2000), these are: (a) lower price and income elasticity of demand for primary products than manufactured goods; (b) technical progress that economizes on the use of primary raw materials in the manufacturing process; (c) technological superiority of developed countries and the control exercised by multinational enterprises based in these countries of the use of sophisticated manufacturing technology; and (d) monopolistic market structure in developed countries, combined with competitive conditions in both commodity and labour markets in developing countries. Compared to today’s ‘high-tech’ time series econometrics, the methodology employed by both Prebisch (1950) and Singer (1950) was very simple in providing the evidence on the declining relative price of primary products to that of manufactured goods for the time period covering the latter half of the nineteenth century to about the ﬁrst half of the twentieth century (Sapsford and Balasubramanyam, 1994).5 By pulling together two sets of overlapping series, Prebisch observed that the NBTT of the United Kingdom for the whole of its merchandise trade had registered a secular improvement during the period from 1876–80 to 1946–47. Since for most of the period under consideration the UK was the world’s most important exporter of manufactures and importer of primary products, Prebisch interpreted this evidence to imply a secular deterioration of the NBTT of primary products traded worldwide. On the other hand, in his descriptive analysis of the problems of specializing in primary production, Singer referred to some statistics reported

3 For the nineteenth century there is some evidence in favour of the classical economists (see Sarkar, 1986). However, since the beginning of the twentieth century, things have changed markedly; it will be shown below that the dominant strand of the recent literature takes the view that there is a fall in the terms of trade for primary commodities. 4 The net barter terms of trade for primary products is deﬁned as the ratio of an index of export prices of primary products to an index of import prices of manufactures. 5 The term ‘high-tech’ time series econometrics comes from Sarkar (1986) and refers to the relatively recent development of unit roots and cointegration techniques and their application to macroeconomic data.

18

Secular Decline in Relative Commodity Prices by the UN to make the point that ‘ . . . the trend of prices has been heavily against sellers of food and raw materials and in favour of sellers of manufactured articles’ (Singer, 1950, p. 477), together with some underlying potential reasons for such a tendency. This is how the Prebisch-Singer (PS) hypothesis came into existence, to be debated for the rest of the twentieth century and beyond. The objections raised against the PS hypothesis, together with the issues explored in the subsequent empirical literature, can be summarized as: (i) the misleading evidence emanating from the inappropriateness of UK terms of trade; (ii) the arbitrariness of the time span; (iii) the use of inadequate data; (iv) the statistical procedure; (v) the omission of other important variables in the analysis; (vi) the failure to take into account improvements in the quality of products; and (vii) the fact that developing countries are not the only exporters of primary commodities (Diakosavvas and Scandizzo, 1991; Sarkar, 1986; and Spraos, 1980). The use of UK terms of trade to draw conclusions about the overall relative price of primary commodities has been discussed at length by Spraos (1980) and Sarkar (1986), who ﬁnd that the choice of the indicator was not unjustiﬁed.6 In the later period, much better data have been used, so this issue is not a major concern. The argument that developing countries are not the only producers and exporters of primary commodities also has no effect on the positive component of the hypothesis, but its implications might be different from those deduced by Prebisch and Singer, i.e. the relative distribution of gains from trade between the ‘centre’ and the ‘periphery’. As regards the criticism of the PS hypothesis that manufactured goods are more subject to quality improvements, potentially making the NBTT of primary goods appear worse, it should be mentioned that there is no measurement of differential qualitative change in the two types of products. Improvements in quality have also taken place in primary commodities and to what extent the prices of manufactured goods reﬂect more of the upward drift on account of quality improvement is not known (Grilli and Yang, 1988; Sarkar, 1986; Spraos, 1980).7 In the following discussion, therefore, this review will attempt to cover the other four issues as they have been explored in a number of important and inﬂuential empirical works. 6 Spraos (1980, p. 113) concludes, ‘ . . . the evidence of Britain’s NBTT to an inference about the relative price of primary products vis-a`-vis manufactures in world-wide trade was not misleading as to direction though it gave an exaggerated impression of the magnitude of deterioration’. On the other hand, Sarkar (1986, p. 361) observes: ‘ . . . Prebisch was to a large extent justiﬁed in choosing the NBTT of Britain . . . as ‘proxy’ for the terms of trade of the industrial region vis-a`-vis the agrarian region of the world’. 7 Grilli and Yang (1988) and Bleaney and Greenaway (1993) cite a few studies that have attempted to measure the effects of changes in the quality of manufactured goods on their prices. However, these studies are unlikely to be representative of the manufacturing goods sector as a whole. On the other hand, there is no study of the impact of quality improvement on primary product prices.

19

The Issue of Declining Commodity Prices

2.2. Empirical Findings to the mid-1980s Between 1950 and 1980 few studies were undertaken to verify the PS hypothesis empirically, although discussions relating to the causes of the change in terms of trade attracted considerable interest.8 In an early attempt Wilson et al. (1969), as reported in Diakosavvas and Scandizzo (1991), considered the NBTT and income terms of trade of developed and least developed countries for the period 1950–65. Taking 1950–53 as the base years, the study found that between 1954–57 and 1962–65 LDCs’ NBTT fell from 98.3 to 90.7. However, it was Spraos (1980) who introduced solid statistical tools into the analysis by empirically estimating the linear trend equations using Singer’s data and its extended version compiled by the author himself.9 He found that the relative price series for the 70-year period up to the outbreak of World War II provided support for the PS hypothesis, although the statistical series used by Prebisch exaggerated the rate of deterioration. However, Spraos observed that the declining rate was not stable and became very weak (‘open to doubt’) if the trend equation was estimated using the dataset extended to 1970. In other words, the ﬁnding was to be seen as the PS hypothesis being subject to the chosen time span. In a subsequent paper, Sapsford (1985) pointed out the problem of structural break in the trend equation estimated by Spraos. The method of trend estimation (or any model) is usually based on the assumption of parameter constancy and the potential problem associated with the stability of the parameters can be tested statistically. There could be several problems with stability of the regression coefﬁcients, viz. only the intercept might change from one sub-sample to another, only the slope parameters might change, or both could change.10 Extending the data series considered by Spraos to 1982, Sapsford’s Chow (1960) test for structural stability supported a once-for-all upward shift in relative price in 1950 without any signiﬁcant change in the declining trend between the pre- and post-war sub-periods. Sarkar (1986) and Thirlwall and Bergevin (1985) have also investigated the differential rates of declining commodity prices for different periods. Earlier in this paper Singer was criticized over his choice of time span, 1876–1938, as the terminal date was marked by the depression of the 1930s which was argued to have been responsible for exaggerating the negative trend. Fitting the trend 8

Diakosavvas and Scandizzo (1991) provide a list of all such studies. The linear trend equation for estimating the growth rate for any variable Y takes the form of lnYt ¼ a þ bT þ et , where ln stands for natural logarithm, a and b are respectively intercept and slope parameters and T is the time trend with, say, 1 for the beginning year of the sample to n, where n is the number of periods under consideration. 10 Consider the equation lnY ¼ a þ bT þ cD þ d(b D)T, where D is a dummy variable indicating a break point in the data and all other variables are deﬁned as above. In estimation if c is signiﬁcant but not d, this will result in intercept shift only. The signiﬁcance of d only will result in shift in slope parameter while the signiﬁcance of b and d will result in changes in both the intercept and slope coefﬁcients. 9

20

Secular Decline in Relative Commodity Prices equation to League and UN series on the NBTT of primary products for the two periods 1876–1929 and 1876–1938, Sarkar observed that both series exhibited a statistically signiﬁcant declining trend and the inclusion of the data for the 1930s only accentuated the existing declining trend.11 Considering the postwar period, Sarkar’s results show that the exclusion of petroleum from the group of primary commodities results in a trend decline rate of 0.89 per cent per annum.12 On the other hand, Thirlwall and Bergevin were interested in the differential rates of decline between the two sub-periods 1954–72 and 1973–82. The trend deterioration in the case of the ﬁrst sub-period was estimated at 1.2 per cent per annum, while the corresponding rate for the latter period was found to be as high as 2.5 per cent per annum.13

2.3. The Grilli-Yang Study and Subsequent Empirical Investigations The empirical work that gave new impetus to the investigation of trends in commodity prices and provided the strongest evidence since the launch of the PS hypothesis is that by Grilli and Yang (1988). The most important contribution of Grilli and Yang was to prepare a consistent dataset. The authors ﬁrst gathered US dollar price indices of 24 internationally traded non-fuel commodities for 1900–86, and then used them to construct an aggregate price index with 1977–79 values of world exports of each commodity used as weights. To obtain the relative price of primary commodities, Grilli and Yang used the UN index of unit values of exports of manufactured goods from industrial countries (MUV) as the deﬂator.14 The original MUV series had two breaks for the years 1915–20 and 1939–47, which the authors ﬁlled in by interpolation. Having constructed the new series of the relative price of primary commodities, estimation of the linear trend equation by OLS resulted in a statistically signiﬁcant trend growth rate of about –0.6 per cent per annum.15 11 The use of data as provided by Lewis (1952) for 1870–1929 yields the lowest declining rate of 0.29 per cent per annum, while the dataset of Schlote (UN, 1949) for 1938–1976 provides the highest rate of 0.84 per cent. 12 This is because of the oil price shock of the 1970s that produced a sharp increase in fuel prices. However, it is now standard practice to consider the NBTT of non-fuel commodities while examining the issue of secular decline. 13 Thirlwall and Bergevin (1985) did not undertake the stability test as suggested by Sapsford (1985). 14 As reported in Grilli and Yang (1988), the estimate of the trend growth rate is not sensitive to the choice of deﬂator. The use of the US manufacturing price index instead of MUV would have produced similar results. 15 Grilli and Yang’s work is not the only attempt in compiling a very long-run series in commodity prices. Earlier, based on data reported in Schlote (1938), W. A. Lewis constructed a long data series for 69 years starting from 1870. The Economist’s ‘index of industrial commodity prices’ uses data since 1862 and is updated regularly. Apart from these, Diakosavvas and Scandizzo (1991) have also attempted to construct a data series, which remains unpublished, for as many as 14 commodities. However, because of the revision of commodity composition

21

The Issue of Declining Commodity Prices Partly because of the availability of consistent and long-time series data as provided by Grilli-Yang, and partly because of the advent of modern time series econometrics of unit roots and cointegration, the period since the late1980s has witnessed a renewed interest in applied works on commodity prices. Studies using the newly developed applied econometric techniques to test the conclusion reached by Prebisch and Singer about the secular decline in commodity prices have become a regular phenomenon. The ﬁrst such notable study was by Cuddington and Urzua (1989); following the development in time series econometrics, they argued that the traditional trend equation estimation for obtaining the long-run growth rate was only valid if the underlying series had been a stationary one.16 If, on the other hand, the variable under consideration has a unit root (i.e. the series is non-stationary), the traditional trend growth equation will have to be modiﬁed.17 This modiﬁcation requires the transformation of a non-stationary series into a stationary one and running a difference stationary (DS) model of the following type as originally proposed by Nelson and Plosser (1982).18 in The Economist’s index and the discontinuation of Lewis’ data series, there are major problems in using them in empirical application. On the other hand, a wider commodity coverage than Diakosavvas and Scandizzo and the systematic use of them in constructing a weighted aggregate series have made the Grilli-Yang dataset the most acceptable. 16 A time series is stationary if its mean, variance and auto-covariance are independent of time. By now there is compelling evidence that many macroeconomic time series are indeed non-stationary, which has some signiﬁcant implications for regression analyses employing OLS. It has been shown that OLS regressions involving non-stationary data might produce not only inconsistent and inefﬁcient estimates but also ‘spurious’ or nonsense relationships. In other words, one could obtain highly signiﬁcant correlation between variables although in reality there might not exist any such relationship. One interesting example of spurious regression is illustrated by Hendry (1980) to show that there has been a strong positive relationship between the inﬂation rate and the accumulated annual rainfall in the United Kingdom! 17 Whether a variable is non-stationary can be determined by testing for the existence of a unit root in its data generation process. The two most popular tests for unit roots are the Dickey Fuller (DF) and Augmented Dickey Fuller (ADF) tests. The DF test is based on the equation: DYt ¼ t þ (c 1)Yt1 þ xT þ et where Y is the variable under consideration, D is the ﬁrst difference operator, subscript t denotes time period, T is the time trend and e is the error term. The null hypothesis for this test is that (c 1) ¼ 0 (i.e. Yt is non-stationary) against the alternative of (c 1) < 0 (i.e. Yt is stationary). The t-test on the estimated coefﬁcient of Yt1 provides the DF test for the presence of a unit root. In the presence of non-stationary variables the distribution of t-test is non-standard and the special critical values for the distribution of the non-standard t-test in the above model have been tabulated by Dickey and Fuller. The ADF test, on the other hand, is a modiﬁcation of the DF, which involves augmenting the DF equation by lagged values of the dependent variables to ensure that the error process in the estimating equation is residually uncorrelated. The null and alternative hypotheses in the ADF equation are the same as the DF regression and so are the critical values. Note that a series without a unit root is also known as a trend stationary process (TSP) while the one with a unit root is a difference stationary process (DS). 18 This is represented by: DlnYt ¼ a þ et , where all variables are deﬁned above and D denotes transformation of the variable into a stationary series. Since the left-hand side in the equation is the proportional growth rate in Y, an estimate of the trend growth rate according to this method is obtained by regressing the growth rate of the relative commodity price against a

22

Secular Decline in Relative Commodity Prices Therefore, for Cuddington and Urzua (1989) the type of equation to be used for the estimation of trend growth rates critically depended on the test for unit root in the relative commodity price series. Visual inspection of the data on relative price of commodity as constructed by Grilli and Yang showed a big spike in 1921, prompting Cuddington and Urzua to consider a once-for-all drop in relative prices for that year in light of which they employed the Perron (1989) test for unit to determine the time series property of the underlying variable.19 The results allowed them to conclude that the relative commodity price series was non-stationary and accordingly they opted for a DS model which yielded a trend rate not signiﬁcantly different from zero. Cuddington and Urzua also employed the Beveridge-Nelson (1981) technique to decompose commodity price movements into permanent and cyclical components and found that roughly 39 per cent of average shock to NBTT was to be viewed as permanent, while the rest was cyclical. Similarly, in another study Cuddington (1992) applied the unit root test to determine the time series property of each of 24 commodity price indices of Grilli and Yang (1988), together with the comparable data for oil and coal. Thirteen commodity price indices appear to be difference stationary process (DSP) while the remainder can be modelled as trend stationary process (TSP). Of the 26 commodities, only ﬁve are found to have a negative trend, while in all other cases the hypothesis of a secular decline in prices is rejected. Like Cuddington and Urzua (1989) and Cuddington (1992), Newbold and Vougas (1996) have applied various univariate time series techniques to determine whether the aggregate relative primary commodity price index is trend or difference stationary. From these tests no conclusive inference could be made about the unit root property of the variable. The authors found evidence for the PS hypothesis when the series was considered to be TSP, but in the case of DSP there was no overwhelming evidence. Despite the ambiguous results derived from the unit root tests, the authors preferred the difference stationary constant, with an error term. This transformation ensures that the residual term is white noise, which otherwise turns out to be non-stationary in the case of the dependent variable possessing a unit root. The transformation from non-stationary to stationary usually requires differencing of the variable. Following Engle and Granger (1987), a variable having unit root on its level but not on its ﬁrst difference is called integrated of order one and is often denoted as e I(1). A second or higher order of differencing might also be required to eliminate the unit root from the data-generating process, although most non-stationary series appear to be e I(1). Non-stationarity of the residuals in the time series regression is considered to be an important problem leading to the potential problem of spurious relationship. On the other hand, even if a regression comprising non-stationary variables yields stationary residuals, the estimated equation may still show a valid long-run relationship. Engle and Granger (1987) show that if two variables, Yt and Xt , are both e I(1), they will have a valid long-run relationship (usually said to be ‘cointegrated’) if residuals from the OLS regression of Xt on Yt are e I(0). 19 It has been shown that investigation of whether a series is TSP or DSP using standard DF and ADF tests can lead to wrong inferences if structural breaks are ignored (Perron, 1989; Zivott and Andrews, 1992).

23

The Issue of Declining Commodity Prices 2.00

Relative Price of Primary Commodities

1.75 1.50 1.25 1.00 0.75 1900

1910

1920

1900

1910

1920

1930

1940

1950

1960

1970

1980

1990

2000

20

0 −20 Changes in the Relative Price of Primary Commodities 1930

1940

1950

1960

1970

1980

1990

2000

Figure 2.1. Grilli-Yang Relative Price of Primary Commodities and its Changes Over Time

model over the trend stationary alternative and concluded the PS hypothesis to be ‘non-proven’. The econometric evidence to nullify the PS hypothesis has been subjected to detailed scrutiny by Sapsford et al. (1992) and Leon and Soto (1997). According to Sapsford et al. (1992), the unit root testing procedure of Cuddington and Uruza was inappropriate, as the equations estimated to ascertain the order of integration of variables contained many insigniﬁcant lagged periods of the dependent variable, exclusion of which would have rejected the null hypothesis of unit root in the data, thereby establishing the superiority of the TS model as against the DS one employed by the authors. Another major problem of Cuddington and Urzua’s ﬁndings is related to the plausibility of the 1920–21 decline in relative commodity price as reﬂected in the Grilli and Yang (1988) dataset. Sapsford et al. argue that the 50.3 per cent fall in the relative price for that particular year may be called into question as the commodity price series constructed by Schlote (1938) reports this fall at 13.5 per cent only. If the decline for 1921 in the Grilli-Yang dataset is replaced by the extent of fall in Schlote (1938), a replication of the Cuddington and Urzua exercise re-establishes a signiﬁcant downward trend in relative price over 1900–83 (Sapsford et al., 1992).

24

Secular Decline in Relative Commodity Prices On the other hand, Leon and Soto (1997) challenge the ﬁndings of Cuddington’s (1992) analysis of a declining rate for 24 individual commodities in the Grilli-Yang dataset. They followed the same approach as Cuddington, but instead of using Perron’s (1989) test, considered the unit root testing procedure of Zivott and Andrews (1992). In Perron’s methodology the test for structural break at a particular time is selected on the basis of data inspection, while Zivott and Andrews’ technique allows for determination of the break point endogenously and statistically. Application of this endogenous break point methodology resulted in 20 commodities (of a total of 24) becoming TSP and to defy the conclusion reached in Cuddington (1992) signiﬁcant and negative trends were observed for as many as 17 commodities. The evidence from decomposition of time series into permanent and cyclical components by Cuddington and Urzua (1989) has also been disputed in subsequent studies. In Ardeni and Wright (1992) a structural time series approach, following Harvey (1989), is undertaken to decompose the aggregate composite commodity price index of Grilli and Yang (1988) into permanent, cyclical and residual components. The authors’ results demonstrate a permanent trend decline in the relative price of commodities at a rate of 0.6 per cent per annum. The experiments in this paper also do not provide support for the 1921 structural break affecting the trend declining rate to any signiﬁcant extent. These ﬁndings are also corroborated by another important study by Reinhart and Wickham (1994). Using the IMF quarterly data on an all non-fuel real commodity price index from the ﬁrst quarter of 1957 to the second quarter of 1993, Reinhart and Wickham ﬁrst tested for unit roots and failed to reject the null hypothesis of non-stationarity, which led them to implement Beveridge and Nelson’s (1981) ARIMA and Harvey’s (1989) structural time series approach to decompose the series into permanent and temporary (or cyclical) components. From the results of both experiments it became clear that the weakness in commodity prices has been permanent in nature. Another study that seems to contradict the PS hypothesis of stable declining terms of trade of primary products is that by Powell (1991). Powell considered cointegration analysis to test for a long-run relationship between a commodity prices index and an index of unit values of manufactures (MUV in the GrilliYang dataset). Both were in nominal dollars and both were found to be nonstationary. Cointegration between the variables, together with the value of the cointegrating parameter being equal to one, would be interpreted as evidence against secular declining terms of trade. The Johansen test for cointegration results showed that the variables were cointegrated with the long-run parameter not statistically signiﬁcantly different from one only when three outliers of 1921, 1938, and 1975 are controlled with a ‘jump term’. From this, Powell concludes that commodity terms of trade are stationary but with three sharp breaks. However, the major problem is that the same results can also be interpreted as a stepwise version of the PS hypothesis with permanent drops

25

The Issue of Declining Commodity Prices in those three years. Besides, although outliers are controlled with a jump dummy, no attempt is made to consider the changes in the cointegration parameter between the outliers. Cashin and McDermott (2002) and Hadass and Williamson (2002) have published two recent studies using different data from those used by Grilli and Yang (1988). Cashin and McDermott (2002) employed The Economist’s index of industrial commodity prices over the period 1862–1999.20 They estimated the trend decline rate in the series to be 1.3 per cent per annum— more than double the estimate made by Grilli and Yang (1988). The local trends (i.e. the trend over a decade) are found to vary remarkably from 2.7 per cent in the 1910s to as high as 6.9 per cent in the 1990s. The authors, however, could not ﬁnd any evidence for a break in the long-run (1862–1999) trend, although the highest possibility of the appearance of such a break occurred in 1917. Setting the sample to 1917–1999 yields a declining rate of 2.3 per cent—much larger than estimated for the full length of the series. On the other hand, Hadass and Williamson (2002) employ a completely different methodology. Unlike the international prices of primary commodities relative to manufactured foods, they gathered the terms of trade data for 1870–1940 in the home markets of 19 countries, which they then divided into the ‘centre’ and the ‘periphery’ using the average unskilled wage or GDP per capita criteria.21 They found that the terms of trade deﬁned as the price of agricultural products relative to that of manufactures improved in every region, which was consistent with their hypothesis of ‘transport revolution’. The main problem with the Hadass and Williamson study is that their sample is limited to only nineteen countries and none of the developing countries in the sample truly reﬂects the typical poor commodity-dependent nation.22 What becomes obvious from the above discussion is that most post-Grilli-Yang studies are plagued with the unit root testing procedure with inconclusive evidence about the exact time series properties of the variable. This problem is essentially inherent in the weak and low power of the unit root testing procedure and as Harris (1995) points out, the most important problem faced when applying the unit root test is their probable poor size and power properties.23 20 The real annual data of The Economist’s index of industrial commodities consist of the nominal industrial commodity price index (dollar-based with base 1845–50 ¼ 100, weighted by the value of developed country imports), deﬂated by the GDP deﬂator of the United States. 21 The authors observed that the share of primary exports in total exports could not be used to deﬁne the centre and periphery as during the sample period primary goods dominated world trade and countries both in the North (Europe and America) and in the South (mainly Asia), as included in the sample, had the same degree of dependence on primary commodities. 22 Developing countries included in the sample are Argentina, Burma, Egypt, India, Korea, Thailand, and Taiwan. 23 The problems of unit root testing procedure have been known for a long time; Engle and Granger (1987) also highlighted the low power of the DF and ADF tests. Considering the strengths and weaknesses of the testing procedures, Gujarati (2003, p. 820) concluded that ‘as yet there is no uniformly powerful test of the unit root hypothesis’.

26

Secular Decline in Relative Commodity Prices This is often reﬂected in the tendency to over-reject the null hypothesis when it is true and under-reject it when it is false. Even studies applying the modern time series techniques to the PS hypothesis are aware of this problem. For example, Newbold and Vougas (1996), having applied all the rigorous techniques in the arsenal of unit roots econometrics, realize that the econometric tests are relatively uninformative on the question of whether the relative price of primary commodities is trend stationary or integrated of order one. It is also clear from the above that whether or not any of the violent ﬂuctuations in the time series of commodity prices has led to a structural break has been the subject of signiﬁcant statistical controversy.24 While the regression methodology is capable of testing for structural breaks, how they affect the unit root property of a variable has not yet been settled in the applied econometrics literature. While the weakness of traditional unit root testing procedures in the presence of structural breaks is supposedly overcome by Perron (1989) or Zivott and Andrews (1992) type tests, none of the procedures can consider more than one structural break in the data. In a recent attempt Kellard and Wohar (2002), employing the Lumsdaine and Papell (1997) methodology for searching two endogenously determined break dates, conﬁrm the trend stationarity of 15 individual commodity prices (out of the 24 in the Grilli-Yang dataset), 12 of which appear to have a declining trend, as opposed to only ﬁve found by Cuddington (1992). Even before wondering at the contrasting evidence, one might ask: why test for only one or two structural breaks in the data and why not more? Therefore, it would not be inappropriate to conclude that, despite the problem of violent ﬂuctuations in the time series of commodity prices, existing econometric procedures are still uninformative in terms of determining how these affect the underlying time series properties. How the variables need to be modelled even when they are integrated has been a matter of careful investigation in econometric theory and applied econometric techniques (e.g. Banerjee et al., 1993; Charemza and Deadman, 1992; Engle and Granger, 1987; Harris, 1995; Hendry, 1995 and 1999). The most important and uncontroversial lesson of this literature is that mere differencing of the variables to transform them into stationary series and using them in OLS regression is tantamount to wiping out long-run information and should be avoided. The suggested procedure is to use some kind of cointegration technique, which makes it possible to obtain both the longrun and short-run estimates of the model. Therefore, even if the time series of relative prices of primary commodities is considered to be non-stationary,

24 For example, Sapsford (1985) considered 1950 to be the year that led to a shift in the intercept of the trend equation, while Cuddington and Urzua (1989) favoured 1921. Powell (1991), on the other hand, introduced a jump term in his regression equation to capture outliers corresponding to 1921, 1938, and 1975.

27

The Issue of Declining Commodity Prices a simple estimate of equation (1) by OLS as implemented in Cuddington and Urzua (1989) and in other studies should be problematic. Only one study (Bleaney and Greenaway, 1993) avoids the problem of unit root testing procedure, yet formulates a more general speciﬁcation of the trend equation that encompasses both trend and difference stationary models.25 The speciﬁcation used by the authors follows an error-correction modelling approach and is thus consistent with a cointegration technique. Updating the Grilli and Yang (1988) aggregate relative prices for primaries, Bleaney and Greenaway’s model provides a trend decline of about 0.84 per cent for the period 1902–91. Since the relative commodity price is found to be unusually high in the earlier part of the twentieth century, to avoid the exaggeration of the downward trend particular emphasis is given to the sample covering 1925–91. This yields a trend growth rate of 0.7 per cent per annum. The results also support a ‘oncefor-all’ drop in the relative prices of primary commodities after 1980.

2.4. Structural Models Low income elasticity of demand, declining intensity of primary resource use in the industrial countries and supply surge in agricultural production are thought to be the most important reasons for the long-run downward trend in real commodity prices.26 However, only a few empirical studies have been undertaken to explain the trend in commodity prices with other economic time series through structural models. As noted in Sapsford and Singer (1998), Borensztein and Reinhart (1994) attempted to explain recent depressed commodity prices by extending the traditional approach to the demand side to include the political and economic transition in Eastern Europe and the former Soviet Union. On the supply side, they emphasized the pressures brought about by the debt crises of the 1980s. Others, however, focused on an explanation of long-term decline in relative prices. Bloch and Sapsford in a number of papers (1992, 1997, 2000) explicitly referred to the explanation advanced by Prebisch (1950) and Singer (1950) with regard to the differences in competitive environment between primary and manufacturing production. In Bloch and Sapford’s models, therefore, wages and prices in primary production are competitively determined, while in the manufacturing sector they are 25

The approach taken by Bleaney and Greenaway (1993) is elaborated in the next section. The increase in supply of agricultural commodities is the result of the entry of new exporters into international markets (e.g. during the 1980s Malaysia and Indonesia became major suppliers of cocoa) as well as of technological progress (e.g. the development and diffusion of fertilizer-pesticides-irrigation mechanisms in crop production). Land under cultivation has also increased in many parts of the developing world along with a sustained increase in yields. For some commodities, the agricultural policies of the industrial countries have also contributed to the rapid expansion in world commodity supplies (Reinhart and Wickham, 1994). 26

28

Secular Decline in Relative Commodity Prices determined by mark-up pricing and union-employer bargaining. Both the level of the mark-up in the manufacturing sector and the wages in either sector may be affected by output levels or by the prices of both types of goods. Estimating the model for the world economy (i.e. using the data on aggregate commodity prices, industrial production, overall manufacturing wages, etc.), the authors ﬁnd some support for the difference in market structure as contributing to the downward trend in the terms of trade.27 However, the main problem of the analysis is that some of the key variables in the model are not statistically signiﬁcant.28 On the other hand, the analytical framework of Deaton and Laroque (2003) makes use of Lewis’s (1954) argument that as long as there is an inﬁnitely elastic supply of labour at the subsistence wage, commodity prices cannot rise and may even decline with local technical progress.29 In this model. commodity supply is assumed to be inﬁnitely price elastic in the long run, and the rate of growth of supply responds to the excess of current price over the long-run supply price. On the other side, demand is related to the level of world income and to the price of the commodity. Deaton and Laroque ﬁtted the model for six commodities over the years 1900–1987.30 The results of the empirical investigation appear to be mixed, with variables of interest in a number of equations failing to become statistically signiﬁcant.

2.5. Concluding Observations From the above review of the literature it may be reasonable to conclude that there is now a broad consensus on the long-term trend deterioration in relative commodity prices. Whilst the trend rate of decline may differ between individual commodities, on the basis of the very long-run data the magnitude of the estimates ranges from 0.6 to 2.30 per cent per annum. There is also some evidence that weakness in prices in the most recent past has been much steeper than the long-run average rate. 27

Their results show that for the period 1948–93 the adverse impacts on the terms of trade of primary products due to a trend difference in wage growth and the trend increase in markups in manufacturing are almost exactly offset by the impact of strong growth of manufacturing production. 28 The model used by Bloch and Sapford is highly aggregative in nature. The authors admit the problem of data, especially with respect to capital stock. The data on wages in the primary sector are proxied by a weighted average of agricultural wages in Mexico, Sri Lanka, India, Chile and Turkey. 29 In his original article, Lewis (1954) considered the price of sugar and real wages of workers in the West Indies. He argued that wages cannot grow because of unlimited supplies of labour at the subsistence wage. Therefore, the beneﬁts of technical progress in sugar production accrued not to workers but to consumers in industrial countries (Deaton, 1999). 30 The implementation of the model requires information on commodity prices, total production of the commodities and world GDP.

29

The Issue of Declining Commodity Prices Nonetheless there are studies where the authors are still sceptical regarding a long-run trend decline. However, the weight of the evidence has certainly led to changes in the position of the World Bank and the IMF with regard to relative commodity prices (Sapsford and Singer, 1998). Until the 1980s, the World Bank and the IMF preferred to take the view that there was price volatility (without a downward trend), despite the existence of statistical evidence on the secular declining trend. However, since the late 1980s, work undertaken by both Bank and Fund economists has conﬁrmed a long-run secular decline in the net barter terms of trade of primary commodities. However, relatively little has been done to explain long-run commodity price behaviour in terms of other factors. Several reasons have been given for the weakness in commodity prices, but robust statistical evidence supporting any of the alternative hypotheses is still unavailable.

30

Table 2.1. Summary of Findings on Secular Decline in Commodity Prices Study

Methodology and Data

Main Finding

Prebisch (1950)

A simple examination of the data by splicing the two partially overlapping series of Schlote (1938; 1952) and the United Nations (1949). The data corresponded to the net barter terms of trade of the UK for the whole of its merchandise trade, the inverse of which could be considered as the terms of trade of primary commodities.

Between the 1870s and 1930s the ratio of prices of primary to manufactured goods fell by 38 percentage points (as shown by Prebisch, Table 1 (p. 9).

Singer (1950)

Descriptive analysis of the problems of specialisation in the primary sector.

No data or statistics have been used for illustration but the author refers to a UN publication to make the point that ‘ . . . the trend of prices has been heavily against sellers of food and raw materials and in favour of the sellers of manufactured articles’ (p. 477 and footnote 4). Also provides reasons for the declining trend in relative prices.

Spraos (1980)

Linear trend equation ﬁtted by the ordinary least squares regression was used to estimate growth rates. Used the dataset as considered by Prebisch (1950) and also compiled a new series to take into account the post-World War II period.

The author found that the balance of evidence from the range of relative price series for the 70-year period up to the outbreak of World War II provided support for the Prebisch-Singer hypothesis, although the statistical series used by Prebisch exaggerated the rate of deterioration ‘at worst by a factor of more than three’ (p. 126). However, if the sample was extended to 1970, the empirical evidence became ‘open to doubt’.

Sapsford (1985)

Linear trend equation corrected for autocorrelation by the CochraneOrcutt iterative method is used to estimate the growth rate. Chow test is carried out to examine the possibility of a structural break between the pre- and post-World War II period as implicit in the ﬁndings of Spraos (1980). The dataset used by Spraos is extended to the early 1980s.

An upward intercept shift in the post-war period is observed, but the shift occurs without any signiﬁcant alteration in the downward trend as between the pre- and post-war sub-periods. The estimated long-run trend growth rate for the period 1900–82 is 1.29 per cent per annum.

Thirlwall and Bergevin (1985)

Trend growth rate estimation for two different sub-periods of 1954–72 and 1973–82. The United Nations quarterly data are used in the analysis.

The trend deterioration for real commodity prices turns out to be 1.2 per cent per annum between 1954 and 1972 while the estimated rate of decline for 1973–82 appears to be more than double at 2.5 per cent per annum.

Sarkar (1986)

Trend growth rate estimation for different periods, pre- and post-World War II, to examine whether Prebish-Singer results are subject to the time span chosen. Data used are taken from League (1945), Lewis (1952), Prebisch (1950), Schlote (1952) and various UN sources and correspond to aggregate price index.

The trend growth rates for 1876 and 1938 range between 0.29 to 0.84 per cent per annum. For the period 1953–80 the trend rate is affected by the inclusion of petroleum in the group of primary commodities because of the oil shock of 1970s. Exclusion of petroleum, however, results in a trend decline rate of 0.89 per cent per annum. (Continued)

Table 2.1. (Continued ) Study

Methodology and Data

Main Finding

Grilli and Yang (1988)

The authors compile US dollar price indices of 24 internationally traded non-fuel commodities for 1900–86. Then an aggregate price index is constructed with 1977–79 values of world exports of each commodity used as weights. The UN index of the unit value of exports of manufactured goods from industrial countries is considered as the deﬂator. The linear trend equation is used to estimate the growth/ decline rate in the aggregate relative price index.

The relative price of non-fuel primary commodities is estimated to have fallen by 0.6 per cent per annum. Signiﬁcant negative trends emerge for most principal commodity sub-groups such as food, non-food agricultural and cereals.

Cuddington and Urzua (1989)

Time series models and the Perron (1988) unit root test are employed to make the distinction between trend stationary and difference stationary processes. In addition, the Beveridge-Nelson (1981) technique is used to decompose price movements into permanent and cyclical components. The Grilli-Yang aggregate index of non-fuel commodity prices (deﬂated by the unit value of exports of manufactured goods from industrial countries) is used in the empirical investigation.

There was a permanent drop in the level of relative primary commodity prices in 1921 but apart from that there is no evidence of secular deterioration. Roughly 39 percent of the average shockto the NBTTcomes out as permanent while the remaining 61 per cent is cyclical and dies out within three years. The permanent component has a one-time drop in 1921 but since then grows at a rate of 0.3 per cent (positive) per year.

Diakosavvas and Scandizzo (1991)

Simple linear and quadratic trend equations are estimated employing the generalised least squares (GLS) procedure. Data on prices of 19 commodities for 1900–82 have been gathered from different sources. The UN index of unit value of exports of manufactured goods from industrial countries has been used as the deﬂator.

For eight commodities, a declining and signiﬁcant trend is discernible, but for six others there is counter evidence.

Powell (1991)

Cointegration analysis undertaken to test for a long-run equilibrium relationship between commodity prices and manufactured goods’ unit values. Used the Grilli-Yang aggregate commodity price index and the index of unit value of manufactured goods from industrial countries.

Controlling for three outliers in 1921, 1938 and 1975, cointegration between commodity and manufactured goods prices is found, with the cointegration parameter being unity. This is then interpreted as the evidence against a ‘stable declining commodity terms of trade’.

Cuddington (1992)

Time series techniques are used to determine whether each of the 24 commodity price indices, as prepared by Grilli and Yang (1988), plus two others on oil and coal contain unit roots or can be modelled as TSP.

Thirteen commodity price indices appear to be DSP while the rest are TSP. Of the 24 individual commodities, only ﬁve are found to have a negative trend as predicted by Prebisch and Singer, while the others have either zero or positive trends leading to the rejection of the secular decline in relative prices of commodities hypothesis.

Sapsford et al. (1992)

The Perron unit root test as in Cuddington and Urzua (1989) is applied, but only with lags that are statistically signiﬁcant to test for the existence of TSP versus DSP data generating process in the relative commodity price index of Grilli and Yang. The data compiled by

The Perron test with low order signiﬁcant lags leads to the rejection of the unit root in Grilli-Yang series as found by Cuddington and Urzua (1989). The secular decline is found to be sensitive to the amount of relative price fall in 1921. If the relative price fall in the Grilli-Yang series is replaced by

Schlote (1952) are used to express scepticism about the massive fall in commodity prices relative to those of industrial goods as reﬂected in aggregate relative price series constructed by Grilli and Yang.

an equal amount of Schlote (1952) dataset, a signiﬁcant downward trend in commodity prices is established for 1900–86.

Ardeni and Wright (1992)

The structural time series approach of Harvey (1989), where the components of the time series are decomposed into the trend, cycle and residuals, and the Grilli-Yang aggregate real commodity price index (updated to 1988) are used.

The estimated trend growth rate is found to be negative and the rate is 0.6 per cent per annum.

Bleaney and Greenaway (1993)

Considered a general error-correction speciﬁcation of the trend equation that encompasses both trend stationary and difference stationary models. For empirical exercise the Grilli-Yang (1988) relative commodity price index is used, updated to 1991.

Since the relative commodity price is unusually high in the earlier part of the twentieth century, to avoid the exaggeration of the downward trend particular emphasis is given to the sample covering 1925–91, in which case the trend rate is estimated to be 0.7 per cent per annum. The evidence of a ‘once-for-all’ drop in the relative prices of primary commodities after 1980 is found.

Reinhart and Wickham (1994)

The ADF, Phillips-Perron and Perron tests are used to check for unit roots and structural breaks in the data. ARIMA and structural approaches are used to decompose the time series into permanent and cyclical components. IMF quarterly data on all non-fuel real (aggregate) commodity price index deﬂated by the IMF index of manufacturing export unit values of industrial countries for 1957:I—1993:II.

Both the ARIMA and structural decomposition techniques present a similar result: the bulk of the price weakness is associated with the secular component and there is no evidence of an abnormally large cycle. Irrespective of the technique used, the downward trend is found to have steepened towards the end of the sample.

Newbold and Vougas (1996)

Univariate time series techniques are used to determine whether the aggregate series of relative prices of primary commodities can be modelled as TSP or DSP. The series of the relative prices of (aggregate) primary commodities under investigation is the one prepared by Grilli and Yang for 1900–87 and subsequently extended by Bleaney and Greenaway (1993).

The evidence of secular decline depends ‘to a substantial degree’ on whether the time series of relative prices is assumed to be trend stationary or integrated of order one for which the authors’ conclusion is that the usual econometric tests are relatively uninformative. In the case of TSP, the best estimate of downward drift is in the neighbourhood of 0.8–0.9 per cent per year, unless the experience of 1921 when there occurred a big fall, is discounted, in which case the ﬁgure falls to about 0.64 per cent. However, if the relative price series is considered to be difference stationary, there is no overwhelming evidence of any downward drift. The authors ﬁnd that the case for trend stationary is not strongly established and therefore their conclusion is that the Prebisch-Singer hypothesis is ‘non-proven’.

Leon and Soto (1997)

Considered a test for ﬁnding structural breaks in the data endogenously as developed by Zivott and Andrews (1992). Used the same dataset of 24 commodity price indices as Grilli and Yang (1988).

In the case of 20 (out of 24) commodities, relative price indices turned out to be TSP. Negative and signiﬁcant trends to support the PS hypothesis were found for 17 commodities.

Kellard and Wohar (2002)

A unit root testing technique developed by Lumsdaine and Papell (1997) that allows for two endogenously determined break dates (unlike the Zivott and Andrews (1992) test that searches for just one) is used to determine the data-generating process. The long-run trend is estimated by adopting ARIMA speciﬁcation. The dataset comprises the same 24 commodities as in Grilli and Yang, but the ﬁgures are updated to 1998.

The tests lead 15 commodity prices to be classiﬁed as trend stationary. In various speciﬁcations with different dummies as required by the unit root test results, only 12 commodities were found to have negative time trend for 50 per cent or more of the time, providing ‘modest’ support for the PS hypothesis. The authors note, ‘ . . . [H]owever this result is sensitive to the decision criterion adopted and one should caution against any quick judgements as to the robustness of the PS hypothesis’ (p. 14). (Continued)

Table 2.1. (Continued ) Study

Methodology and Data

Main Finding

Cashin and McDermott (2002)

The Economist’s index of industrial commodity prices covering the period 1862–1999 is used. The trend growth rate is estimated for three sub-periods to examine whether there has been any change in the trend rate.

There has been a downward trend in real commodity prices of about 1.3 per cent per year over the past 140 years. Although not statistically signiﬁcant, the highest possibility of structural break is detected in 1917. The average annual rate of decline between 1971 and 1999 is estimated to be 2.3 per cent. No support for a break in the long-run trend decline in commodity prices.

Hadass and Williamson (2002)

A completely different methodology is used. Data on the terms of trade in home markets for a number of 19 sample countries between 1870 and 1940 are gathered. The sample countries are then divided into ‘centre’ and ‘periphery’ using such indicators as the unskilled real wage and GDP per capita criteria.

The terms of trade are found to have improved in every region during the sample period, which is explained by the ‘revolution’ in the transport sector. In fact, for the period 1870–1940, the terms of trade are found to have improved more in the periphery than in the centre. ‘However, consistent with Singer’s prediction, these positive relative price shocks had an asymmetric impact in centre and periphery, boosting growth in the centre and suppressing it in the periphery’ (p. 22).

3 Long-Run Trend in the Relative Price: Empirical Estimation for Individual Commodities Mohammad A. Razzaque, Philip Osafa-Kwaako, and Roman Grynberg

In this chapter we estimate the trend growth rate in relative prices for individual commodities. Most studies consider the aggregate or composite relative price index in order to examine the validity of the PS hypothesis. However, individual commodity prices rather than the composite price index are more important for countries in ascertaining their problems or prospects related to export earnings and balance of payments emanating from trends in commodity prices. Some commodities might be subject to much steeper declining rates than the overall relative price index, in which case the movement in the aggregate price index would hardly reveal the practical consequences for countries specializing in them. In fact, the Prebisch–Singer thesis can also be considered for each of the major commodity groups (such as food, agricultural raw materials, minerals, etc.) and for the individual products comprising the broad classiﬁcations. It might also be of interest to see whether the hypothesis holds for all commodities, and if not, whether some characteristic features can be identiﬁed for the commodities that do not experience deteriorating net barter terms of trade. Most importantly, any general policy conclusion can only be deduced if a similar trend is revealed for most individual commodities. The empirical results, as provided in Bleaney and Greenaway (1993), show that different broad categories of primary commodities appear to exhibit price behaviour which is different from the aggregate relative price index. If this is so, then the examination of price behaviour at the individual level should be the most appropriate way of evaluating trends in commodity prices.

35

The Issue of Declining Commodity Prices

3.1. Methodology The literature review in the previous section included studies investigating individual commodities, usually with data taken from the work of Grilli and Yang (1988) or its updated version. However, these studies (e.g. Cuddington, 1992; Kellard and Wohar, 2002; and Leon and Soto, 1997) place too much emphasis on testing unit root properties in order to determine the appropriateness of trend stationary vis-a`-vis difference stationary models for estimating the trend equation. The results of these studies are highly inﬂuenced by whether the relative price data (for any individual commodity) are to be considered as TSP or DSP (Cuddington, 1992) and how many break points are being explicitly tested for in the process of determining the time series property of the variable under consideration (Leon and Soto, 1997; Kellard and Wohar, 2002). The underlying econometric tests have low power, as well as methodological issues that are as yet unsettled.1 Therefore, using these estimation techniques is unlikely to be informative. One alternative to avoiding the unit root testing procedure and the ensuing pitfalls is to follow the methodology used by Bleaney and Greenaway (1993) by constructing a general error correction model that encompasses both the trend and difference stationary models. Instead of prior testing of time series properties of the data, this methodology aims at minimizing the possibility of uncovering a spurious trend by appropriately allowing for possible dynamics involved in the determination of the trend rate. Despite the standard practice in modern applied time series econometrics of testing for integrating orders of variables before running a regression, the use of such a framework that does not require prior testing for unit roots may be appropriate, given very recent developments in the ﬁeld. In fact, Pesaran et al. (2001) have devised a new approach to testing for the existence of a valid long-run relationship between variables which is applicable irrespective of whether the underlying variables are stationary, integrated of order 1 or mutually cointegrated. It has been argued that using this procedure it is unnecessary to establish the order of integration of the variables prior to estimation of the long-run relationship and that therefore, unlike typical applications of cointegration analysis, this method is not subject to the well-known shortcomings associated with the pre-testing techniques. The recent development thus supports the methodology employed by Bleaney and Greenaway (1993), especially when it has been demonstrated that the determination of unit root properties for commodity prices series with violent ﬂuctuations is anything but straightforward. Further, the Bleaney-Greenaway approach happens to be a special case in the Pesaran et al. framework. In the following we outline the methodology 1

For example, how to choose a break point in unit root testing procedure or how many breaks are to be considered.

36

Long-Run Trend in the Relative Price adopted by Bleaney and Greenaway (1993) and relate this to the framework of Pesaran et al. (2001). Consider the standard trend equation: lnRP ¼ a þ bt þ u

(1)

where RP is the relative price and all other variables are as deﬁned in the previous section. According to Cuddington and Urzua (and all others follow them), equation (1) can only be employed if lnRP is trend stationary. If lnRP is non-stationary and is ~I(1), the relevant model to be estimated is: D lnRP ¼ b þ u

(2)

Instead of using (1) or (2), Bleaney and Greenaway started with an autoregressive model with a time trend included: lnRP ¼ a þ bt þ clnRPt1 þ u

(3)

The main difference between (1) and (3) is the inclusion of a lagged dependent variable as a regressor. Equation (3) can be rearranged to obtain: D lnRP ¼ a þ bt þ c lnRPt1 þ u

(4)

where, c ¼ c 1. Equation (4) becomes an ideal error-correction model if c is negative, statistically signiﬁcant and greater than 1, (i.e. 1 < c < 0). In that case, the change in lnRP is negatively related to its current level and this will pull back the short-run deviations to the steady state long-run trend path. By contrast, if c ¼ 0, lnRP may be considered as a random walk with increasing variance over time. In essence, an error-correction representation in (2) is only possible if the prices of primary products and manufactured goods are cointegrated.2 In the estimation of (4), if b 6¼ 0, and c < 0, lnRP has a non-zero deterministic trend, i.e. it has a long-run tendency to revert to a non-zero trend following any short-term disturbances. The combination of b ¼ 0 and c ¼ 0 will imply no long-term trend of lnRP but the series tends to be pulled back towards its historical mean. Thus both ‘b < 0 and c ¼ 0’ and ‘b<0 and c < 0’ will provide empirical support for the deteriorating trend hypothesis. For two other combination possibilities, viz. b ¼ 0 with c ¼ 0 and b 6¼ 0 with c ¼ 0, the results should be interpreted as the evidence for the series to be a random walk with zero mean and random walk with a drift respectively.3 2 This result is due to Engle and Granger (1987). Therefore, Bleaney and Greenaway (1993) suggest that equation (4) should be considered as an alternative to test for cointegration between primary product and industrial goods prices, as is done in Powell (1991) and noted in the previous section. 3 For a series that is a random walk with zero mean, its history gives no indication of its future path. In future its value can be greater or less than its current value. For a random walk

37

The Issue of Declining Commodity Prices Estimation of equation (4), therefore, does not require the testing of the variables for unit roots a priori. Let us now brieﬂy consider the Pesaran et al. (2001) methodology that makes the testing for time series properties redundant and provides further justiﬁcation for the use of (4) in our empirical estimation. The procedure suggested by Pesaran et al. is based on an OLS estimation of an unrestricted error correction model, a general speciﬁcation of which with respect to two variables, X and Z—whose integrating orders are not determined a priori but are expected to be either zero (i.e. trend stationary) or 1 (i.e. ﬁrst difference stationary)—and the trend term, T, can be written as:

DlnXt ¼ a þ uT þ g ln Xt1 þ jZt1 þ

p X i¼1

pi D ln Xt1 þ

g X

di DZti þ «i

(5)

i¼0

Estimation of (5) in itself is not interesting because the existence of a long-run relationship can only be tested by examining the joint null hypothesis that g ¼ j ¼ 0 with the help of either a Wald or an F-test. The presence of a long-run relationship requires the rejection of this null. However, as the asymptotic distribution of these statistics is non-standard, Pesaran et al. provide the necessary critical upper (FU ) and lower (FL ) bound for the F-test.4 The FU statistics are derived under the assumption that all variables are I(1) and the FL values consider all of them to be I(0). If the computed F statistic (F), which is obtained by restricting that g ¼ j ¼ 0, is greater than the critical upper value, i.e. F > FU , the null is to be rejected and a valid long-run relationship among the variables may be ascertained. If F < FL , then no long-run relationship exists; ﬁnally, if FL < F < FU , the test is inconclusive. Pesaran et al. (p. 290) clearly point out that ‘[I]f the computed Wald or F-statistic falls outside the critical value bounds, a conclusive inference can be drawn without needing to know the integration/cointegration status of the underlying regressors.’5 From (5) it is observed that if there is no other explanatory variable (apart from the trend term), the Pesaran et al. speciﬁcation becomes the standard Dickey-Fuller unit root testing equation—just as the one used by Bleaney and Greenay (1993). Under such a circumstance, the statistical signiﬁcance of the lagged level dependent variable will be regarded as a proof of the long-run relationship. However, if the dependent is non-stationary on its level, the

with drift, if the estimated b is positive, it is more probable that it will be greater than its current value in the future and the opposite is true if b turns out to be negative. 4 Pesaran et al. give both the critical values for Wald and F-statistics. In this paper we will only consider the F-statistics. 5 In equation (5) pi and di give the short-run estimates of the parameters. The long-run parameter values can be obtained by noting that there is no change in hthe isteady state such that: DXt ¼ DZt ¼ 0. This would imply the long-run coefﬁcient on Z as: gj .

38

Long-Run Trend in the Relative Price distribution of T-statistics is non-standard and Pesaran et al. suggest that the critical value for testing the statistical signiﬁcance of the lagged level dependent variable in the absence of any other explanatory variable will correspond to Dickey and Fuller’s (1979) unit root T-statistics.6 Therefore, an error-correction type trend equation model that encompasses both trend and difference stationary models such as the one in (4) not only avoids the problems of unit root testing procedures but is also justiﬁed. In Dickey-Fuller type equations, such as the one in (4), special importance is given to the problem of serial correlation. The concern over the presence of serial correlation is usually addressed by the inclusion of one or more lags of the dependent variable as regressor.7 Thus a more general form of equation (4) can be written as:

DlnRP ¼ a þ bT þ

m X

h DlnRPt1 þ FlnRPtm þ ut

(6)

i¼1

X where, F ¼ I hi And the long-run trend rate is given by: b ¼ F1 :

3.2. Estimation Results We now turn to the results. Except for one instance, the data used here are for prices of individual commodities. Two different datasets have been used to obtain the information on prices. First, an attempt was made to gather the data on individual commodities in Grilli and Yang (1988). Of the 24 commodities, data was obtained on 13 covering the period 1900–87.8 These are cocoa, coffee, tea, bananas, sugar, rice, wheat, maize, cotton, jute, palm oil, copper and tin.9

6 This, in effect, implies that in the absence of any other explanatory variables (apart from the constant and trend term) the statistical signiﬁcance of the lagged level dependent variable is to be considered as evidence for a valid long-run relationship irrespective of the unit root property of the data. 7 A general practice in the case of annual data is to include at least one lag of the dependent variable and then to check for the residual autocorrelation problem. For quarterly data at least four lags are used. 8 We thank Angus Deaton for providing us with the Grilli-Yang data on commodity prices for these 13 commodities. From an e-mail communication, it was learnt that the World Bank no longer has access to the information on the individual commodities price series used in the Grilli-Yang study. 9 The commodities for which information could not be obtained were aluminium, beef, hides, lamb, lead, rubber, silver, timber, tobacco, wool and zinc.

39

The Issue of Declining Commodity Prices The series was then updated to 2001 using comparable information.10 All data were gathered in nominal US dollars and then the unit value index of the manufactured goods exports of the industrial countries was used as the deﬂator to compute commodity-speciﬁc net barter terms of trade.11 Apart from the Grilli-Yang dataset, the UNCTAD database on commodity prices was used to estimate the trend growth rate for as many as 60 individual commodities.12 The longest span of the data available from UNCTAD is 1960–2002. In most cases these data were available in US dollars and the unit value index of manufactured goods exports of developed market economy countries was used as the deﬂator.

3.2.1. Trend growth rates of relative prices for commodities in the Grilli-Yang dataset Figure 3.1 plots the updated 13 commodity-speciﬁc relative prices in the GrilliYang dataset. All relative prices exhibit wide ﬂuctuations with spectacular peaks and troughs. Nevertheless, even a casual look at the graph clearly reveals a declining trend in the net barter terms of trade of rice, wheat, maize, cotton and palm oil. For bananas a strong declining trend is discernible from around 1930 and for tea and jute from the mid-1950s. Apart from two skyscrapers, a deteriorating trend in the real price of sugar is also clear. Tin is the only commodity that witnessed a strong positive trend until the early 1970s, largely because of the success of the International Tin Agreement (ITA). Since then, the real tin price began falling before the major crash of the mid-1980s, which coincided with the collapse of the ITA. The most striking feature of Figure 3.1 is that since the 1970s a strong downward trend in the real prices of all commodities is apparent. Table 3.1 provides the regression results for the commodities in Figure 3.1. It needs to be mentioned here that except lnRPt1 , for all variables the standard t-ratios are valid, which implies that as a rule of thumb if the t-ratio is greater than 2 the respective coefﬁcient is statistically signiﬁcantly

10 Apart from jute, price series for commodities were updated using the information in various issues of Global Economic Prospects, published by the World Bank. Price data in the International Financial Statistics Yearbook of the IMF were used to build the series on jute for 1987–2001. 11 Note that Grilli and Yang (1988) used MUV as the deﬂator. For later periods we use what UNCTAD now publishes as the unit value index of manufactured goods exports from the developed market economy countries. Appendix 1 gives the graphical plots of these two series, which show that the series are almost the same. A linear trend line ﬁtted through the scatter of the two series resulted in a R2 value of 0.999 with the coefﬁcient on the explanatory variable very close to one (the restriction that the coefﬁcient was exactly one could not be rejected at the 1 per cent error probability level). 12 These data were accessed from the Commodity Price Bulletin of UNCTAD. For this study the online version of the dataset was used from the website: www.unctad.org

40

Long-Run Trend in the Relative Price 1.5

Cocoa 1.0

Coffee

2.0

1.0

1.5

0.5

1.0

Tea

2.0

Bananas

1.5 0.5

1.0

0.5 1900

1950

2000 1900

1950

2000 1900 3

Sugar

4

2

Rice

1950

2000 1900

Wheat 3

1950

2000

Maize

2 2

2

1 1

1900

1950

2000 1900

1950

3

Cotton

2000 1900 4

Jute

2

3

2

1 1950

Palm Oil

1900 1.00

1950

2000 1900

2000 1900

2000

Copper

1.0

1 1950

2.0

1950

1.5

2 1

1

2000 1900

1950

2000 1900

1950

2000

Tin

0.75 0.50 0.25 1900

1950

2000

Figure 3.1. Relative Prices of 13 Commodities: 1900–2001 Note: The ﬁgures correspond to relative commodity prices.

different from zero at the 5 per cent error probability level. For lnRPt1, however, the estimated t-ratios should be compared with those of the critical values computed by Dickey and Fuller (1979) to draw inferences. These critical values are considerably higher than the standard t-ratios. In fact, in order for the ln RPt1 term to be statistically signiﬁcantly different from zero, the computed t-ratio should be as high as 3.13 (absolutely) at the 10 per cent signiﬁcance level. Following the usual practice with Dickey-Fuller regressions, the ﬁrst order lagged dependent variable (i.e. ˜lnRPt1 ) is always retained in the equation irrespective of its statistical signiﬁcance. In only a few cases additional lags were also included to remove the problem of serial correlation. In a number of equations, regression residuals turn out to be non-normal, which should be considered as a serious problem preventing the drawing of valid inferences. As sudden and precipitous price ﬂuctuations are common, as reﬂected in Figure 3.1, it is unsurprising that a simple trend equation will fail to explain such movements, resulting in residuals that are not normally distributed. Bleaney and Greenaway (1993) also encountered the problem of

41

The Issue of Declining Commodity Prices non-normality in estimating the trend growth rate in the aggregate commodity price index for which they re-estimated their equation after dropping the ﬁrst 25 years of data from their sample, arguing that those years were associated with exceptionally violent movements of commodity prices. Figure 3.1, however, does not seem to suggest that at the individual commodity levels the movement in commodity prices prior to 1925 was different from that in the latter period and therefore it was decided not to curtail the sample to tackle the problem of non-normality. Instead, we have used dummy variables to control for the sudden rise(s) and decline(s) in commodity prices. This approach is tantamount to pulling the atypical data points to a normal year, which is deﬁned by the trend equation. All dummies inserted in all equations were found to be highly signiﬁcant. The estimated equations with the dummies can be considered as the preferred speciﬁcation and growth rates corresponding to these equations will used for reference. Results reported in Table 3.1 show that for ten out of thirteen commodities the estimated coefﬁcients on the trend equation are negative; only for cocoa, coffee and tin is the sign on the coefﬁcients positive. For eight commodities— tea, sugar, rice, wheat, maize, cotton, jute and palm oil—the estimated trend is negative and statistically signiﬁcant at the 10 per cent conﬁdence level. Among the three commodities with a positive sign, only the trend rate for tin is signiﬁcant over the period 1900–87. In all regressions the lagged level dependent variable (lnRPt1 ) is negative and less than zero, as is expected in the case of an error-correction model. In as many as eight cases the T-ratio on lnRPt1 is higher than the Dickey-Fuller critical value (at least at the 10 per cent level), which implies that for these commodities a valid long-run trend growth rate can be estimated irrespective of the order of integration of the real price series. Although for another ﬁve commodities the estimated T-ratio on the lagged level term is lower than the critical value, it is always signiﬁcantly different from zero, considering the standard test of signiﬁcance for stationary variables. Indeed, if any of these relative price series is TSP, estimation of trend growth rate for it from the regression results can be considered valid.13 In the column for ‘trend rate’, the long-term trend growth rate in relative price (in per cent per annum) has been computed for all commodities for which the coefﬁcient on the trend term appears to be signiﬁcant at least at the 10 per cent error probability level. For cocoa, coffee, banana, copper and tin the trend term is not signiﬁcant and the exact interpretation will depend on whether one considers lnRPt1 in those equations to be signiﬁcant or not.14 13

Note that since there is no pre-testing for unit root, it is not known a priori whether any of these ﬁve series are TSP. In some studies, when pre-testing for unit roots is not done, the standard T-ratios are used to make inferences about the statistical signiﬁcance of the lagged dependent variable (e.g. Athukorala, 2000). 14 As mentioned above, if the statistical signiﬁcance of lnRPt1 is to be determined on the basis of the Dickey-Fuller critical values, then the variable is signiﬁcant only in the case of copper.

42

Long-Run Trend in the Relative Price If the coefﬁcient on the lagged level variable is to be considered signiﬁcant, the real price series of these commodities have no long-term trend but they tend to be pulled back towards their historical mean.15 Negative trend growth rates have been estimated for tea, sugar, rice, wheat, maize, cotton, jute, palm oil and copper. The trend rates lie between 0.79 and 1.43 per cent per annum and the results show that during the past century most commodity prices have fallen at an annual rate of above 1 per cent. This is considerably higher than the estimates of Grilli and Yang (1988) and Bleaney and Greenaway (1993) which were in the range 0.6 to 0.7 per cent per annum.

3.2.2. Estimation for commodities in the UNCTAD database 3.2.2.1. ESTIMATES FOR BROAD COMMODITY GROUPS The commodity price bulletin of UNCTAD provides information on prices for a large number of individual commodities since 1960. It also provides an aggregate commodity price index and price indices for another four broad commodity groups, viz. food and beverages, vegetable oils and oilseeds, agricultural raw materials, and minerals and metals. Before analysing the individual commodities, Table 3.2 gives an estimate for the broad commodity groups. In general, the estimation of the trend equation was affected by the normality problem mainly due to the sudden jump in commodity prices around the mid-1970s, as Figure 3.2 exhibits one clear peak for all broad commodity groups. Therefore, for most equations dummy variables were included to control for these sharp price movements. As with the previous cases, the equations with the dummies are considered to be the preferred speciﬁcations. The results reported in Table 3.2 show that for every broad commodity group the trend variable appears to be statistically signiﬁcant. In every preferred speciﬁcation, apart from the one for the food and beverage group, the computed T-ratio on the lagged dependent variable (lnRPt1 ) exceeds the DickeyFuller critical values at least at the 90 per cent level.16 Although a ﬁrm conclusion about the long-run relationship cannot be made for food and beverages, separate estimates for the ‘food only’ and ‘beverages only’ sub-groups strongly rejected the null hypothesis of statistical insigniﬁcance of lnRPt1 , suggesting that irrespective of the order of integration of the dependent variables the estimated trend growth rates are valid. In no regression is lnRPt1 insigniﬁcant in comparison with the t-statistics following standard distribution and applicable for drawing inferences in the case of stationary variables.

15

Otherwise, the relative price series form a random walk with zero mean. The equation for food and beverages did not show any residual non-normality problem and therefore no dummy variable was inserted to control for the sharp rise in 1973, as shown in Figure 3.2. 16

43

Table 3.1. Regression Results for 13 Commodities (with Updated Grilli-Yang Series: 1900–2001) DlnRPt Cocoa

Constant

T

lnRPt1

DlnRPt1

0.13* (1.76)

0.00038 (0.044)

0.114 (2.15)

0.15 (1.53)

0.75** (2.98)

.00093 (0.11)

0.86 (1.76)

0.10 (1.13)

Coffee Tea

0.16** (1.90) 0.04 (1.21) 0.92*** (0.3)

0.0014 (0.15) 0.001* (1.68) 0.001** (2.15)

0.19 (2.94) 0.11 (2.24) 0.09 (2.50)

0.05 (0.55) 0.07 (0.1) 0.04 (0.48)

Banana Sugar

0.05*** (3.11) 0.29*** (3.29) 2.82*** (5.19)

0.004 (1.16) 0.004*** (3.02) 0.004*** (4.97)

0.14*** (3.29) 0.40*** (4.84) 0.38*** (6.57)

0.10 (1.02) 0.17* (1.69) 0.16** (2.38)

0.003*** (3.62) 0.003** (3.81) 0.0037*** (4.19) 0.003*** (4.59) 0.004*** (3.80) 0.0032*** (3.46) 0.0025*** (3.94) 0.0017** (2.11) 0.0012* (1.62) 0.0036*** (3.52) 0.003*** (3.11) 0.004 (0.75) 0.0007 (1.02) 0.001 (1.6)

0.25** (4.07) 0.24*** (4.21) 0.32*** (4.84) 0.32*** (5.10) 0.35*** (4.41) 0.22* (3.23) 0.196** (3.89) 0.19** (3.26) 0.16** (3.85) 0.29*** (4.16) 0.26** (3.97) 0.18* (3.21) 0.20** (3.39) 0.16 (2.76)

0.29*** (2.98) 0.29 (3.31) 0.28*** (2.89) 0.27*** (2.95) 0.09 (0.93) 0.084 (0.33) 0.09 (0.76) 0.06 (0.10) 0.05 (0.58) 0.21** (2.10) 0.14 (1.47) 0.11 (1.11) 0.10*** (0.09) 0.153 (1.66)

Rice

0.19*** (3.48) 0.47*** (2.16) Wheat 0.27*** (4.34) 0.82*** (5.23) Maize 0.31*** (3.92) 1.77*** (5.41) Cotton 0.16*** (3.80) Jute 0.11** (2.20) 0.77*** (3.65) Palm Oil 0.18*** (3.11) 0.60*** (2.85) Copper 0.016 (0.49) Tin 0.17** (2.47) 0.76 (4.64)

DlnRPt2 0.28** (2.82) 0.29*** (3.09) — — —

Dummies

Adjusted Serial Functional Heterosce- Trend R2 Correlation Form Normality dasticity rate (%)

None

0.125

0.30

0.60

8.96**

0.06

D47

0.23

0.03

1.14

2.66

0.66

None None D85, D77, D84, D54 — None — None — D20, D21, D63, D65, D74, D80 — None — D73, D82 — None — 0.54*** (3.74) — None — D21, D38, D48 — None — None — D86 — None — D86 — None None 0.032 (0.09) D86

0.06 0.02 0.33

0.02 4.15** 0.72

0.48 0.60 2.38

4.38 11.23** 2.50

0.099 0.37 1.47

1.04 1.26

0.05 0.17 0.64

2.95 2.50 3.22

0.03 79.69** 1.86

3.91 0.64 0.41

White 1.02 3.03

1.02 1.21

0.15 0.34 0.23 0.28 0.15 0.40 0.08 0.07 0.21 0.13 0.24 0.07 0.08 0.19

3.23 0.017 0.09 0.94 1.33 4.40 3.67 1.07 0.17 5.19 1.58 1.94 5.65*** 2.89

0.91 0.57 17.41*** 0.10 1.88 0.006 0.31 1.04 0.39 8.70*** 7.88*** 0.39 3.24 2.04

0.05 0.77 0.18 0.69 0.66 1.08 White 0.99 0.48 3.08 0.22 0.35 0.12 0.48

1.25 1.28 0.92 1.17 1.18 1.43 1.29 0.9 0.79 1.25 1.17

23.97*** 2.04 1.38 5.90* 21.88*** 4.89 0.29 14.67*** 1.13 6.33** 1.00 2.51 9.17*** 2.98

Note : Figures within the parentheses indicate t-ratios. Statistical signiﬁcance at the 1, 5, and 10 per cent levels is indicated by ***, **, and * respectively. Critical values for the coefﬁcient of lnRPt1 at the 10, 5 and 1 per cent signiﬁcance levels are, respectively, 3.13, 3.45 and 4.10. Variables with the letter ‘D’ followed by two digits indicate a dummy variable. For example, D73 indicates a dummy variable with 0 for 1973 and 1 for all other years. All dummies are inserted separately and are always signiﬁcant at the 1 per cent level. ‘White’ indicates that due to heteroscedasticity standard errors are derived from the White’s (1980) heteroscedasticity consistent variance-covariance matrix. implies that the coefﬁcient on the trend term is not signiﬁcant and hence the trend growth rate is not estimated and can be considered to be zero.

Table 3.2. Regression Results for Broad Commodity Groups as in UNCTAD Commodity Price Bulletin: Annual Data (1960–2002)

(DlnRPt )

Constant

Aggregate 0.19** (2.68) Commodity Price Index 0.09*** (5.96) Food and 0.18** (2.34) Beverages - Food only 0.30*** (3.03) 1.79*** (8.14) -Beverages only Vegetable Oils and Oilseeds

0.05 (0.79)

T

0.007*** (2.82) 0.356 (3.06)

DlnRPt1

DlnRPt2 —

None

0.14

1.49

1.15

4.53*

1.77

1.96

0.007*** (3.71) 0.419*** (4.53) 0.025 (0.19) 0.007** (2.45) 0.30 (2.78) 0.16 (1.05)

— —

D73, D74 None

0.48 0.11

0.85 1.72

1.33 1.05

1.15 3.09

0.27 0.53

1.82 2.37

0.008*** (2.82) 0.40*** (3.51) 0.009*** (4.57) 0.51*** (6.53)

0.32** (2.18) 0.15 (1.41)

— —

0.20 0.65

0.94 0.16

0.19 0.96

8.23*** 2.05

0.08 1.53

2.19 1.85

0.006* (1.94)

0.26** (2.42)

0.14 (0.89)

—

None D80, D73, D74 None

0.07

0.002

0.21

9.30***

0.009

2.63

0.28* (3.33) 0.302 (2.05)

0.002 (0.02) 0.17 (1.16)

—

D76, D77 None

0.42 0.35

0.78 0.61

0.30 5.79**

4.21 1.09

0.42 1.09

2.25 2.92

0.01** (2.84) 0.005** (2.56)

0.49** (3.79) 0.43*** (2.90)

0.31** (2.07) 0.37*** (2.75)

— —

D73 None

0.29 0.37

0.04 0.088

4.42 1.99

0.28 0.04

2.38 1.38

0.331** (2.16) 0.01*** (3.83)

0.295*** (3.14) 0.32*** (3.84) 0.59*** (4.19) 0.24* (1.66)

D73, D76 None

0.76 0.27

0.95 0.05

0.37 5.28

1.29 1.21

1.08 1.75

1.27*** (5.01) 0.006** (2.29) 0.16 (1.57) 0.008* (1.96)

0.70*** (3.48) Agricultural 0.13** (2.17) Raw Materials 0.64*** (8.22) Minerals 0.27*** (3.74) and Metals

lnRPt1

Trend rate Serial Functional HeterosceAdjusted 2 (per cent) Correlation Form Normality dasticity Dummies R

0.18 (1.15)

0.02 77.86***

1.00 0.67

Note: Critical values for the coefﬁcient of lnRPt1 at the 10, 5 and 1 per cent signiﬁcance levels are, respectively, 3.13, 3.45 and 4.10. Variables with the letter ‘D’ followed by two digits indicate a dummy variable. For example, D73 indicates a dummy variable with 0 for 1973 and 1 for all other years. Figures within the parentheses indicate t-ratios. ‘White’ indicates that due to heteroscedasticity standard errors are derived from the White’s (1980) heteroscedasticity consistent variance-covariance matrix. implies that the coefﬁcient on the trend term is not signiﬁcant and hence the trend growth rate is not estimated and can be considered to be zero.

The Issue of Declining Commodity Prices Aggregate Commodity Price Index 2.0

3

1.5

2

1.0

1 1960

4

Food and Beverages

1970

1980

1990

1960

2000 2.0

Food Only

3

1.5

2

1.0

1

1970

1980

1990

2000

1980

1990

2000

Beverages Only

0.5 1960

1970

2.0

1980

1990

2000

1960

1.5

1.50 Vegetable Oils and Oilseeds 1.25

1.0

1.00

1970

Agricultural Raw Materials

0.75

0.5 1960

1970

1.50

1980

1990

2000

1960

1970

1980

1990

2000

Minerals, Ores, and Metals

1.25 1.00 0.75 1960

1970

1980

1990

2000

Figure 3.2. Real Prices of Broad Commodity Groups Note: All prices are relative to the unit value of the index of manufactured goods exports of developed market economy countries. Source : Authors’ calculation based on data from Commodity Price Bulletin (UNCTAD).

Vegetable Oils Food and and Oilseeds Beverages

Beverages

Food

Aggregate Minerals, Ores Agricultural Relative Price and Metals Raw Materials

0

per cent per annum

−0.5

−1

−1.5

−2

−2.5

Figure 3.3. Estimated Growth in Relative Prices for Broad Commodity Groups Source : Based on the results associated with the preferred speciﬁcations in Table 3.2.

46

Long-Run Trend in the Relative Price According to the estimates in Table 3.2, trend growth rates for prices of broad commodity groups fell by between 1.08 to 2.92 per cent per annum. The rate of decline during the past 40 years was lowest for agricultural raw materials and highest for vegetable oils and oilseeds and food and beverages (Figure 3.3). On the whole, the aggregate relative price for primary commodities has been subject to an annual trend deterioration of 1.82 per cent. It is important to note that our preferred speciﬁcation does not exaggerate the rate of decline in commodity prices. In fact, in every case the growth rate associated with the preferred speciﬁcation in Table 3.2 is lower than the equations that do not include any dummy variable to control for residual non-normality. Most dummy variables used to control for a sharp rise in prices fall within the relatively early years of the sample, resulting in a negative effect on the magnitude of the trend growth rate. 3.2.2.2. ESTIMATES FOR INDIVIDUAL COMMODITIES Within each of the broad commodity classiﬁcations, it is possible to estimate the growth rate in relative price for several individual commodities. UNCTAD’s Commodity Price Bulletin provides a wealth of information on prices of many commodities which are narrowly deﬁned, and data on them are gathered in a consistent manner. Subject to the availability of data for a reasonable time period, 17 individual food and beverages products, 9 vegetable oils and oilseeds, 17 agricultural raw materials and 17 commodities in the minerals, ores and metals sub-group were used for empirical estimation. The unit value index of manufactured goods exports of developed market economy countries was used to convert the nominal price series in real terms. Despite frequent ﬂuctuations, a close look at the graphical plots of the real prices of the individual commodities, as given in Appendix 3.10–3.12, reveals a declining trend in real prices for most of the commodities. Initial experiments with the estimation of the trend equation revealed problems related to the model diagnostic tests in a number of equations. The main source of the problems could be found to be associated with the commodity price boom of the mid-1970s. Therefore, for some commodities our preferred speciﬁcation includes dummy variables to control for atypical price rises. One important feature of the speciﬁcation is that if the dummies were not included, growth rates for real prices of commodities would have been higher (absolutely). Therefore, the estimated models, as presented in Appendices 3.1–3.4, do not exaggerate the trend decline in commodity prices. Among the seventeen products in the food and beverages category, the sign on the trend term for all commodities except for white pepper turns out to be negative (see Appendix 3.1).17 Only for cocoa and white pepper is the trend 17 Among the 17 commodities, four types of coffee and two types of wheat are included. Appendices 3.13 and 3.14 show that prices of different varieties of the two products move quite closely.

47

The Issue of Declining Commodity Prices term found to be not statistically different from zero. The lagged level dependent variable in every equation is correctly signed and is always signiﬁcant in comparison with the standard t-statistics. When compared with the Dickey-Fuller critical values, the statistical signiﬁcance of the variable is retained for all commodities except coffee and beef. The estimated trend growth rate varies between 3.27 per cent per annum for tea and 0.92 per cent for bananas. All nine vegetable oil and oilseed commodities have a signiﬁcant negative trend growth rate along with the statistical signiﬁcance of lnRPt1 (Appendix 3.2). Turning to agricultural raw materials, Appendix 3.3 only shows a signiﬁcant positive trend rate for wood items such as non-coniferous woods, sawn wood, tropical logs and plywood. While the estimated trend in the cases of jute and sisal failed to become statistically signiﬁcant, for cotton (various types), linseed oil, leaf tobacco, cattle hides and rubber there was evidence of signiﬁcant declining terms of trade. Among the seventeen products covered in the category of minerals, ores and metals, as many as thirteen have a signiﬁcant lag dependent variable. However, the coefﬁcient on the trend term in the equations for phosphate rock, nickel (cathodes), reﬁned lead, tin (ex-smelter), gold and zinc are not signiﬁcant. On the whole, the application of the error-correction model in the estimation of trend rate is therefore found to be satisfactory. The correct sign and the signiﬁcance of the error-correction term even after comparing with the Dickey-Fuller critical values in the overwhelming majority of the equations suggest that the trend growth rates are valid without a priori knowledge about the order of integration of the variables. Figure 3.4 summarizes the trend growth rates in relative prices over 1960–2002 by individual commodities.18 It is found that two minerals, tungsten ore and silver, have witnessed the steepest decline over the past four decades. The trend declining rates for tea and coffee are found to be much higher than estimates using the very long-term data of 1900–2001.19 Among the cereals, the real price decline for rice has been much worse than that for wheat and maize. For eight commodities, cocoa, sugar, white pepper, jute, phosphate rock, tin, gold and zinc, the trend growth rates are not statistically signiﬁcantly different from zero. While the results for cocoa in both the very long-term sample and the sample beginning from 1960 are qualitatively the same, for tin the positive rate of growth for 1900–2001 has now been turned into one of no signiﬁcant trend. The results in Figure 3.4 cannot be readily compared with those of Table 3.1, which uses very long time-series data. Given the substantial ﬂuctuation in commodity prices, the estimation of the trend growth rate will be affected by

18 In the case of different varieties, a simple average of the estimated growth rates has been used. 19 For tea, the long-term trend growth over the period of 1900–2001 was found to be 1.25 per cent per annum, while for coffee no signiﬁcant rate could be found.

48

Cocoa Sugar White Pepper Jute Phosphate Rock Tin Gold Zinc Tropical Logs Plywood Sawn wood Non-Coniferous Wood

Long-Run Trend in the Relative Price 3

2

0

Tea Rice Coconut Oil Coffee Palm Kernel Oil Copra in Bulk Cotton Palm Oil Cattle Hides Rubber Cotton Seed Oil Beef Yellow Maize Yellow Soybean Copper Soybean Meal Crude Soybean Oil Maize Wheat Sunflower Oil Linseed Oil Groundnut Oil Leaf Tobacco Aluminium Lead Fish Meal Banana Iron Ore Nickel Manganese Ore Sisal

growth rate (per cent per annum)

1

−1

−2

−3

−6

Tungsten Ore

−5

Silver

−4

Figure 3.4. Trend (1960–2002) Growth Rates for Individual Commodities with UNCTAD Data

the time span chosen for analysis. The review of the literature in the previous section also highlights this problem. While a very long-term analysis, such as the one covering 100 years, is useful in understanding the evolution of price movements and in studying the pattern and nature of mean reversion in the data, trends emanating from a relatively recent past are probably more informative in understanding the implications for developing countries. There is not much point in arguing about whether to make the starting point of the sample 1940, 1960, or 1970; nevertheless it might be useful to study the trend in the post-war period. However, while 1970 should be avoided as the initial point because of the commodity price boom, a starting point in the 1980s reduces the number of observations that can be worked with.20 On the other 20 Moreover, in the 1980s commodities prices were already very low. Maizels (1992) shows that relative prices in the 1980s were lower than those during the great depression of the 1930s. Bleaney and Greenaway (1993) report a 37 per cent downward jump in commodity prices after 1980 compared to the average for 1925–1991.

49

Maize

Cotton

Rice

Tea

Sugar

Palm oil

Wheat

Jute

Copper

Banana

Coffee

Cocoa

The Issue of Declining Commodity Prices

0

GY 1960–2002

GY 1900–2001

UNCTAD 1960–2002

Figure 3.5. Trend Growth Rates since the 1960s: Grilli-Yang versus UNCTAD Data

hand, the data show that in the 1960s most relative commodity prices were quite stable; therefore, the starting point of the estimates presented should not be inappropriate. One serious concern is whether these results should be considered as evidence for a potential structural break in the very long-run trend equation of 1900–2001. The issue of structural break has been discussed in a number of studies, including Bleaney and Greenaway (1993), where the authors found statistical support for a once-for-all drop in commodity prices after the 1980s. Thus the possibility of a structural break in the very long-run trend equation cannot be overlooked. However, exactly what time frame should be considered for testing such a break will remain an important issue to be resolved if such a debate is to be informative.21 It is also true that the precise point of structural break might be different for different commodities. What will be the magnitude of the trend decline rate in real commodity prices if the Grilli-Yang type long-run data series is restricted to one comparable with the time frame of the UNCTAD databases? To answer this question trend growth rates for 13 commodities using the 1900–2001 dataset were also estimated for a period from 1960 to the end of the sample. In Figure 3.5 the results are compared with those reported in Table 3.1, together with those 21 In the previous section it was found that using the aggregate relative price of primary commodities, Cuddington and Urzua (1989), Sapsford (1985) and Powell (1991) found structural breaks in different years.

50

Long-Run Trend in the Relative Price plotted in Figure 3.4. It now becomes obvious that the estimates from the UNCTAD data and from the Grilli-Yang data for the comparable period beginning in 1960 do not provide very different results. The biggest discrepancy between the two series is for jute. This is because in the UNCTAD data the trend rate for jute appears to be not signiﬁcant, while using the updated GrilliYang data, results in the trend term are signiﬁcant only at a somewhat lower level of statistical signiﬁcance (i.e. at the 10 per cent level). Figure 3.5 also shows that, with the exception of sugar, jute and cocoa, the growth rate over 1900–2001 is much lower than the sample comprising the data for only the post-1960 period.

3.3. Conclusion The empirical evidence presented in this chapter strongly shows the presence of a statistically signiﬁcant declining trend in the relative price of most individual commodities. When the data spanning the very long period of 1900–2001 are considered, the estimated trend rates lie between 0.79 and 1.43 per cent per annum. Much higher rates of decline are observed for the relatively recent period. Between 1960 and 2002 the aggregate relative price of commodities has fallen at an annual rate of 1.82 per cent, with the corresponding ﬁgures for individual commodities ranging from 0.9 to 3.50 per cent. Therefore, the use of very long time-series data considerably undermines the magnitude of the deterioration of relative commodity prices during the recent past.

51

Appendix 3.1. Estimated Trends in Relative Prices for 17 Food and Beverage Products (With UNCTAD Data 1960–2002) Adjusted R2

Normality

Heteroscedasticity

4.06

2.41

0.02

3.07

1960–2002

0.65

1.23

0.06

2.11

1960–2002

0.33

CochraneOrcutt 1.66

2.66

1.90

0.16

3.45

1960–2002

D76

0.30

2.24

2.16

2.19

0.25

2.60

1960–2002

D76

0.28

0.03

0.81

0.54

2.01

1960–2002

—

D77, D84

0.63

3.58

1.68

2.01

0.68

3.27

1960–2002

—

D90, D74

0.44

0.001

0.06

1.33

1.15

1.98

1960–2002

—

D73

0.46

0.33

0.41

5.29*

0.16

1.78

1960–2002

—

D73, D90

0.35

0.22

2.38

0.44

0.55

2.00

1960–1997

—

D73

0.47

3.14

1.45

1.55

0.73

2.98

1960–2002

—

D74, D80

0.64

2.46

0.11

0.92

0.38

1960–2002

—

None

0.11

0.64

5.91**

2.52

2.52

2.22

1960–2002

—

D73

0.39

0.70

0.025

0.43

0.13

2.19

1960–2002

—

None

0.19

1.23

0.02

0.11

0.55

0.92

1960–2000

—

D97

0.63

0.007

2.85

1.81

0.011

1960–2002

0.33*** (3.40) —

D73

0.72

1.07

3.42

1.83

0.29

2.08

1960–2002

D73

0.46

1.46

1.79

1.55

2.15

1.02

1960–2002

Constant

T

lnRPt1

DlnRPt1

DlnRPt2

Coffee-1

0.92*** (3.10) 0.87*** (3.86) 0.82*** (3.42) 0.79*** (3.34) 0.64*** (2.88) 1.49*** (7.46) 0.12 (0.58) 0.65*** (5.11) 0.17 (0.92) 1.06*** (5.78) 2.76*** (7.47) 0.13** (2.11) 0.62*** (4.66) 0.05 (1.40) 0.32* (1.61) 1.11*** (7.23) 1.15*** (5.35)

0.0089** (2.18) 0.0053* (1.94) 0.008** (2.37) 0.0084** (2.10) 0.0043 (1.31) 0.022*** (4.53) 0.0085*** (2.92) 0.0077*** (2.82) 0.008** (2.48) 0.0138*** (3.51) 0.0039 (1.05) 0.0065** (2.44) 0.010*** (3.23) 0.005** (2.45) 0.002 (0.85) 0.0105*** (3.49) 0.007*** (2.38)

0.29 (2.64) 0.25 (2.88) 0.24 (2.97) 0.325 (2.97) 0.26* (3.11) 0.69*** (5.25) 0.43*** (4.30) 0.43** (4.08) 0.41** (3.60) 0.46*** (4.14) 0.47*** (5.74) 0.29 (2.81) 0.466** (3.98) 0.54* (3.31) 0.33*** (5.08) 0.51*** (4.17) 0.71*** (4.97)

0.086 (0.54) 0.30* (1.69) 0.30** (2.06) 0.27* (1.73) 0.45*** (2.98) 0.168 (1.34) 0.34** (2.64) 0.36** (2.79) 0.31 (2.10) 0.29** (2.27) 0.27*** (2.88) 0.14 (0.90) 0.33** (2.48) 0.13 (0.80) 0.69*** (6.92) 0.03 (0.26) 0.18 (1.35)

—

D77

0.23

1.12

—

D76

0.31

—

D76

0.21 (1.44) —

Coffee-2 Coffee-3 Coffee-4 Cocoa Tea Wheat, Argentina Wheat, US Maize Rice Sugar Beef Yellow Maize Bananas White Pepper Soybean Meal Fish Meal

Dummies

Serial Correlation

DlnRPt

Functional Form

Trend rate (per cent)

Sample

Appendix 3.2. Trend Growth Rates in Relative Prices for 9 Vegetable Oils and Oilseed Products (With UNCTAD Data 1960–2002) DlnRPt

Constant

T

lnRPt1

DlnRPt1

Yellow Soybean Crude Soybean Oil Sunﬂower Oil Groundnut Oil Copra in bulk Coconut Oil

0.81*** (6.10) 0.70** (2.26) 0.19 (0.67) 0.65*** (3.19) 1.09*** (3.50) 1.85*** (4.50) 1.81*** (4.14) 1.24*** (3.69) 0.70*** (3.73)

0.009*** (3.34) 0.0101** (2.73) 0.006* (1.89) 0.008** (2.54) 0.024*** (4.07) 0.02*** (4.29) 0.02*** (3.74) 0.013*** (3.03) 0.10*** (2.86)

0.42*** (4.15) 0.49*** (4.31) 0.37*** (3.44) 0.56*** (4.13) 0.86*** (5.16) 0.80*** (5.33) 0.74*** (4.88) 0.54** (4.09) 0.47** (3.68)

0.08 (0.68) 0.21 (1.63) 0.12 (0.95) 0.18 (1.28) 0.24** (1.55) 0.15* (1.06) 0.10 (0.67) 0.1 (0.97) 0.002 (0.018)

Palm Kernel Oil Palm Oil Cottonseed Oil

Adjusted R2

Serial Correlation

Functional Form

Normality

Heteroscedasticity

Trend rate (per cent)

D73

0.49

0.05

0.25

1.46

0.93

2.18

1960–2002

D73, D74, D86 D74, D86

0.54

2.61

0.61

1.80

1.58

2.04

1960–2002

0.43

0.27

0.26

1.72

2.34

1.86

1960–2002

D74

0.31

0.30

0.09

0.25

1.05

1.55

1960–2002

—

D74

0.41

3.73

3.58

0.018

0.06

2.74

1960–2002

—

D74, D84

0.47

3.02

4.81

1.29

0.016

2.93

1960–2002

—

D74, D84

0.43

2.98

3.01

0.44

0.05

2.75

1960–2002

D74, D84

0.33

3.81

4.36

2.43

0.22

2.55

1960–2002

D74

0.14

3.37

2.11

2.16

0.0003

2.39

1960–2002

Dln RPt2 —

—

Dummies

Sample

Appendix 3.3. Estimated Trends in Relative Prices of 16 Products in Agricultural Raw Materials (With UNCTAD Data 1960–2002) Adjusted R2

Serial Correlation

Functional Form

Heteroscedasticity

Trend rate (per cent)

None

0.15

1.25

0.003

4.36

2.31

2.29

—

None

0.15

0.79

0.001

1.65

2.06

2.10

—

None

0.29

3.35

0.41

2.13

2.26

3.50

0.21 (0.99)

—

None

0.29

0.002

0.58

2.52

0.80

2.81

1960– 2002

0.27 (2.05) 0.41* (3.56)

0.03 (0.29) 0.41** (2.72)

—

D84, D86

0.69

0.23

0.09

1.24

0.85

—

None

0.23

0.14

0.008

1.20

0.04

1960– 2002 1960– 2002

0.47** (3.65) 0.41 (2.28) 0.58** (3.34) 0.80*** (4.55) 0.48 (2.84) 0.76*** (5.59) 0.76*** (5.59) 0.57*** (4.81) 0.51** (3.71) 0.64*** (4.39)

0.36*** (2.32) 0.15 (0.74) 0.29 (1.60) 0.24 (1.50) 0.166 (0.87) 0.17 (1.62) 0.17 (1.62) 0.47*** (3.72) 0.36** (2.26) 0.40** (2.62)

—

None

0.22

0.08

0.71

1.27

1.36

1.23

—

None

0.33

0.26

7.37**

0.47

White

þ1.88

—

None

0.21

0.11

0.002

0.02

0.10

þ1.87

—

None

0.39

0.05

1.05

5.47

2.83

0.69

—

None

0.15

0.36

0.88

0.002

2.13

—

None

0.70

0.001

0.05

0.51

0.77

þ1.35

—

D73, D93

0.70

0.002

0.05

0.51

0.77

þ1.35

—

D74

0.51

0.002

3.23

0.72

0.01

1.79

—

None

0.22

3.70

1.81

3.38

0.55

1.23

—

None

0.29

0.44

0.05

0.28

0.03

2.48

DlnRPt

Constant

T

lnRPt1

DlnRPt1

DlnRPt2

Cotton (US, Memphis) Cotton (US, New Orleans) Cotton (Outlook Index A) Cotton (Outlook Index B) Jute

0.16** (2.03) 0.16** (2.08) 0.73** (2.77)

0.088** (2.52) 0.0088** (2.56) 0.026*** (2.95)

0.38 (2.56) 0.42 (2.66) 0.76* (3.14)

0.06 (0.36) 0.028 (0.16) 0.07 (0.35)

—

0.71** (2.49)

0.024** (2.65)

0.85* (3.31)

0.20 (0.82) 0.06 (0.97)

0.009 (1.51) 0.0037 (1.31)

0.15 (1.76) 0.18** (2.42) 0.26*** (2.12) 0.24*** (3.13) 0.006 (0.10) 0.85 (0.16) 0.85*** (5.17) 0.79*** (3.90) 0.10** (2.15) 0.27** (2.70)

0.005** (1.79) 0.007** (2.77) 0.109** (2.46) 0.005** (2.31) 0.0024 (0.97) 0.10*** (4.15) 0.010*** (4.15) 0.103** (2.73) 0.006** (2.85) 0.015*** (3.38)

Sisal (Tanzania/ Kenya) Sisal (Uganda) Non-coniferous wood Sawn Wood Tropical Logs Tropical Logs (Gabon) Plywood/sheet Plywood/cubic metre Linseed Oil Leaf Tobacco Cattle Hides

Dummies

Normality

Sample 1960– 2002 1960– 2002 1960– 2002

1960– 2002 1972– 2002 1970– 2002 1970– 2002 1970– 2002 1963– 2002 1963– 2002 1960– 2003 1963– 2003 1962– 2002

Rubber in bales Phosphate rock Manganese ore Iron ore Tungsten ore Copper, Grade A Copper, Wire Brass Nickel, LME Nickel, Cathodes Lead, LME Reﬁned Lead Aluminium, high grade Tin, LME Tin, Malaysia Gold Silver Zinc, Special Zinc, Prime Western

0.45*** (3.42) 1.04*** (9.77) 0.11*** (2.51) 0.64 (5.84) 0.19 (1.80) 0.36*** (3.17) 0.23** (2.83) 0.99*** (4.96) 0.84*** (5.04) 0.33*** (2.93) 0.1977 (1.90) 0.29*** (3.42) 0.05 (1.32) 0.04 (0.55) 0.58*** (4.04) 0.26 (1.65) 0.06 (0.93) 0.0077 (0.16)

0.016*** (3.48) 0.0003 (0.02) 0.002* (1.66) 0.003* (1.76) 0.009** (2.14) 0.011** (3.07) 0.007** (2.76) 0.007* (1.71) 0.003 (1.56) 0.009** (2.74) 0.002 (0.99) 0.0094*** (3.22) 0.005** (2.08) 0.006 (1.85) 0.002 (0.98) 0.011** (2.0) 0.0022 (0.87) 0.002 (1.06)

0.66*** (4.25) 0.27*** (4.61) 0.41*** (4.91) 0.33** (3.52) 0.18 (2.38) 0.43** (3.47) 0.40* (3.15) 0.58* (3.40) 0.53*** (4.07) 0.42* (3.27) 0.31 (2.32) 0.80*** (4.47) 0.10 (1.12) 0.14 (1.28) 0.29*** (4.49) 0.29** (2.59) 0.54*** (4.10) 0.60** (3.94)

0.29** (1.87) 0.27*** (4.61) 0.66*** (5.52) 0.26** (2.26) 0.23 (1.46) 0.16 (1.04) 0.12 (0.77) 0.22 (1.51) 0.22* (1.78) 0.13 (0.85) 0.04 (0.28) 0.28** (1.74) 0.14 (0.91) 0.01 (0.07) 0.10 (0.78) 0.20 (1.13) 0.35*** (2.32) 0.28* (1.81)

—

None

0.28

0.76

2.07

2.14

0.07

2.45

—

D74

0.79

0.05

0.008

2.98

0.42

—

None

0.49

1.28

2.40

1.72

2.09

0.63

—

D75, D82

0.60

0.73

1.44

2.23

1.27

0.83

—

None

0.10

0.35

0.24

1.15

0.16

4.96

—

None

0.19

0.17

5.34

0.57

2.02

2.60

—

None

0.15

0.29

0.40

1.03

0.16

1.77

—

D88

0.58

1.11

0.001

2.37

0.45

1.30

—

D88

0.56

0.76

0.002

1.76

0.50

—

None

0.17

3.69

0.54

4.11

0.74

2.29

—

None

0.07

2.42

0.01

0.64

0.18

—

None

0.31

0.04

0.82

6.01**

0.34

1.18

—

None

0.08

0.06

0.013

14.24***

0.77

§

—

None

0.05

0.97

1.26

70.02***

3.18

§

—

D80

0.58

1.49

0.34

2.17

0.43

—

None

0.13

0.03

0.89

6.07**

0.04

3.96

—

None

0.26

0.15

0.09

11.06***

0.51

—

None

0.25

0.88

0.24

0.56

Note : Zinc, Special: The use of dummy for 1973 to control non-normality of errors did not make the coefﬁcient on the trend variable signiﬁcant.

4.30

1960– 2002 1960– 2002 1960– 2002 1960– 2002 1960– 2002 1960– 2002 1960– 2002 1970– 2002 1960– 2002 1960– 2002 1960– 2002 1960– 2002 1960– 2002 1960– 2002 1970– 2002 1970– 2002 1960– 2002 1960– 2002

Appendix 3.4: Estimated Trends in Relative Prices of 17 Products in Minerals, Ores, and Metals (With UNCTAD Data 1960–2002) DlnRPt

Phosphate rock

Constant

1.04*** (9.77) Manganese ore 0.11*** (2.51) Iron ore 0.64 (5.84) Tungsten ore 0.19 (1.80) Copper, Grade A 0.36*** (3.17) Copper, Wire Brass 0.23** (2.83) Nickel, LME 0.99*** (4.96) Nickel, Cathodes 0.84*** (5.04) Lead, LME 0.33*** (2.93) Reﬁned Lead 0.1977 (1.90) Aluminium , high grade 0.29*** (3.42) Tin, LME 0.05 (1.32) Tin, Malaysia 0.04 (0.55) Gold 0.58*** (4.04) Silver 0.26 (1.65) Zinc, Special 0.06 (0.93) Zinc, Prime Western 0.0077 (0.16)

T

lnRPt1

DlnRPt1

0.0003 (0.02) 0.002* (1.66) 0.003* (1.76) 0.009** (2.14) 0.011** (3.07) 0.007** (2.76) 0.007* (1.71) 0.003 (1.56) 0.009** (2.74) 0.002 (0.99) 0.0094*** (3.22) 0.005** (2.08) 0.006 (1.85) 0.002 (0.98) 0.011** (2.0) 0.0022 (0.87) 0.002 (1.06)

0.27*** (4.61) 0.41*** (4.91) 0.33** (3.52) 0.18 (2.38) 0.43** (3.47) 0.40* (3.15) 0.58* (3.40) 0.53*** (4.07) 0.42* (3.27) 0.31 (2.32) 0.80*** (4.47) 0.10 (1.12) 0.14 (1.28) 0.29*** (4.49) 0.29** (2.59) 0.54*** (4.10) 0.60** (3.94)

0.27*** (4.61) 0.66*** (5.52) 0.26** (2.26) 0.23 (1.46) 0.16 (1.04) 0.12 (0.77) 0.22 (1.51) 0.22* (1.78) 0.13 (0.85) 0.04 (0.28) 0.28** (1.74) 0.14 (0.91) 0.01 (0.07) 0.10 (0.78) 0.20 (1.13) 0.35*** (2.32) 0.28* (1.81)

Dln Dummies Adjusted Serial Functional Normality Heterosced- Trend rate RPt2 R2 Correlation Form asticity (per cent)

Sample

–

D74

0.79

0.05

0.008

2.98

0.42

1960–2002

–

None

0.49

1.28

2.40

1.72

2.09

0.63

1960–2002

–

D75, D82

0.60

0.73

1.44

2.23

1.27

0.83

1960–2002

–

None

0.10

0.35

0.24

1.15

0.16

4.96

1960–2002

–

None

0.19

0.17

5.34

0.57

2.02

2.60

1960–2002

–

None

0.15

0.29

0.40

1.03

0.16

1.77

1960–2002

–

D88

0.58

1.11

0.001

2.37

0.45

1.30

1970–2002

–

D88

0.56

0.76

0.002

1.76

0.50

1960–2002

–

None

0.17

3.69

0.54

4.11

0.74

2.29

1960–2002

–

None

0.07

2.42

0.01

0.64

0.18

1960–2002

–

None

0.31

0.04

0.82

6.01**

0.34

1.18

1960–2002

–

None

0.08

0.06

0.013

14.24*** 0.77

§

1960–2002

–

None

0.05

0.97

1.26

70.02*** 3.18

§

1960–2002

–

D80

0.58

1.49

0.34

2.17

0.43

1970–2002

–

None

0.13

0.03

0.89

6.07**

0.04

3.96

1970–2002

–

None

0.26

0.15

0.09

1960–2002

–

None

0.25

0.88

0.24

1960–2002

11.06*** 0.51 4.30

0.56

Note: Zinc, Special: The use of Dummy for 1973 to control of non-normality of errors did not make the coefﬁcient on the trend variable signiﬁcant.

Long-Run Trend in the Relative Price Appendix 3.5. Description for Food-Commodities used from UNCTAD Commodity Price Bulletin Name Coffee-1 Coffee-2 Coffee-3 Coffee-4 Cocoa Tea Wheat, Argentina Wheat, US Maize Yellow Maize Rice Sugar Beef Bananas Pepper

Product Description Coffee, Brazilian and other natural Arabicas, ex-dock NY (¢/lb.) Coffee, other mild Arabicas, ex-dock NY (¢/lb.) Coffee, Robustas, ex-dock NY (¢/lb.) Coffee, composite indicator price 1976 (¢/lb.) Cocoa, average daily prices NY/London (¢/lb.) All teas, London auction prices Wheat, Argentina, Trigo Pan Upriver, f.o.b. Wheat, US, n8 2, Hard Red Winter, f.o.b. Gulf ports Maize, Argentina, c.i.f. Rotterdam Yellow maize, n8 3, US, c.i.f. Rotterdam White milled rice, 5% broken, Thailand, f.o.b. Bangkok Sugar in bulk, Caribbean ports, f.o.b. (I.S.A.) (¢/lb.) Frozen and boneless beef (mainly Australia), US ports (¢/lb.) Fresh bananas, Central America and Ecuador, f.o.b. US ports (¢/lb.) White pepper, 100% Sarawak, Singapore, closing quotations

Appendix 3.6. Description of Vegetable Oils and Oilseeds used from UNCTAD Commodity Price Bulletin Name Soybean Meal Fish Meal Yellow Soybean Crude Soybean Oil Sunﬂower Oil Ground Nut Oil Cora in bulk Coconut Oil Palm Kernel Oil Palm Oil Cottonseed Oil

Product Description Soybean meal 44/45%, Hamburg f.o.b. ex-mill Fish meal 64/65%, any origin, candf, Hamburg Yellow soybeans, n8 2, US, c.i.f. Rotterdam Crude soybean oil, Dutch, f.o.b. ex-mill Sunﬂower oil, E.U., f.o.b. N.W. European ports Groundnut oil, any origin, c.i.f. Rotterdam Copra in bulk, Philippines/Indonesia, c.i.f. European ports Coconut oil, Philippines/Indonesia c.i.f. N.W. European ports Palm kernel oil, Malaysia, c.i.f. Rotterdam Palm oil, 5% ffa, Indonesia/Malaysia, c.i.f., N.W European ports Cottonseed oil, PBSY, US, f.o.b. Gulf ports

Appendix 3.7. Description of Agricultural Raw Materials used from UNCTAD Commodity Price Bulletin Commodity Name Cotton, US, Memphis Cotton, US, Orleans Cotton Outlook Index A Cotton Outlook Index B Jute Sisal, Tanzania/Kenya Sisal, Uganda Non-coniferous Wood Tropical Logs Tropical Logs, Gabon Sawn Wood Plywood, sheet Plywood, metre Linseed Oil Leaf Tobacco Hides Rubber in Bale

Product Description Cotton, US Memphis/Eastern, Midd.1–3/32’’, c.i.f. (¢/lb.) Cotton, US Orleans/Texas, Midd.1-1/32’’, c.i.f. (¢/lb.) Cotton Outlook Index A (M 1–3/32’’) (¢/lb.) Cotton Outlook Index B (coarse count) (¢/lb.) Jute BWD, Bangladesh, f.o.b. Mongla Sisal, n8 3, long, Tanzania/Kenya, c.i.f. London Sisal UG, East Africa, c.i.f. London Non-coniferous woods, UK Import price index ($ equivalent) [1995¼100] Tropical logs, Sapelli LM, UK import price, f.o.b. ($/m3) Tropical logs, Okoume, LM, f.o.b. Gabon ($/m3) Sawn wood, Dark Red Meranti, Malaysia, select and better, c.i.f. French ports ($/m3) Plywood, S.E. Asian Lauan, 4mm, wholesale price, Tokyo (¢/sheet) Plywood, S.E. Asian Lauan, 4mm, wholesale price, Tokyo ($/m3) Linseed oil, any origin, ex-tank, Rotterdam Leaf tobacco, US import unit value Cattle hides, suspension dried, 8/12 lb. Tanzania ($/100kg) Rubber in bales, Singapore n81 RSS, f.o.b.

57

The Issue of Declining Commodity Prices Appendix 3.8. Description of Minerals, Ores, and Metals used from UNCTAD Commodity Price Bulletin Name Phosphate rock Manganese ore Iron ore Tungsten ore Aluminium Copper Copper wire bars Nickel, LME Nickel, cathodes Lead, LME Reﬁned Lead Zinc, Special high grade Zinc, Prime Western Tin, LME Tin, Malaysia Gold Silver

Product Description Phosphate rock, 70% BPL, Khouribga, f.a.s. Casablanca Manganese ore, 48/50% Mn, c.i.f. Europe Iron ore, Brazilian to Europe, 64.5% Fe, f.o.b. (¢/Fe unit) Tungsten ore, Wo3 > 65%, c.i.f. UK ($/t.Wo3) Aluminium high grade, LME, cash Copper, grade A, LME, cash Copper, wire bars, US producer, f.o.b. reﬁnery (¢/lb.) Nickel, LME, cash Nickel cathodes, New York dealer (¢/lb.) Lead, LME, cash settlement ($/t) Reﬁned lead, North America producer price (¢/lb.) Zinc, special high grade, LME, cash settlement Zinc, Prime Western, delivered, North America (¢/lb.) Tin, LME, cash Tin, ex-smelter price, Kuala Lumpur Gold, 99.5% ﬁne, afternoon ﬁxing London ($/troy ounce) Silver, 99.9%, Handy and Harman, New York (¢/troy ounce)

140

MUVGY

UNCTAD

120

1985=100

100

80

60

40

20

1984

1982

1980

1978

1976

1974

1972

1970

1968

1966

1964

1962

1960

0

Appendix 3.9. MUV of Grilli-Yang Dataset and UNCTAD Unit Value of Exports of Manufactured Goods from Developed Market Economy Countries

58

Coffee, composite (1976)

2.5

Tea

Cocoa

2.0

Wheat, Argentina

2.0

2

2.0

Wheat, US, red hard

2.0 1.5

1.5

1.0

1.0

0.5

0.5

1.5

1.5 1

1.0

1.0 1960 2.0

1980

1960

2000 4

Maize, Argentina

1980

2000

2

1.0

1980

2000

1960

1980

2000

Fresh Bananas

1980

2000

0.5 0.75 1960

1980

1960

2000 3

Crude Soybean Oil

2.0

Beef (boneless)

1960 1980 2.0 Yellow Maize

2000

1.5 1.0

1960 4

1980

2000

3

Soybean Meal

1960

1980

2

1

1 1960

2.0

2

1.5

1

1.0

1980

3

2.0

Fish Meal

1980

2000

Yellow Soy beans

1.5 1.0 1960

2000

Groundnut Oil

1960

2000 2.5

3

2

2000

Sunflower Oil

1980

4

White Pepper

1.00

2.5

2000

0.5 1960

1.0

1980

1.0

5

1 1.25

1960 1.5

10

3

1.5

1960 Sugar

Rice

1980

2000

1960 3

Copra in Bulk

2

2

1

1

1980 Coconut Oil

2000

1.5 1.0

0.5

0.5 1960 3

1980

2000

Palm Kernel Oil 2 1

1960 2.0

1980

2000

Palm Oil

1.5

1.5

1.0

1.0

59

1980

2000

1980

2000

Cottonseed Oil

0.5

0.5 1960

1960 2.0

1960

1980

2000

1960

Appendix 3.10. Real Prices of 23 Food and Beverages Items

1980

2000

1960

1980

2000

1960

1980

2000

60

1.5

1.0

1.5

2.0

Cotton, US, Orleans

Cotton, US, Memphis 1.5

1.0 1960 1980 Cotton Outlook, Index B

2000

Cotton Outlook, Index A

1.5 1.0 1960

1980

2000 3

Jute

2

1960 1980 Sisal, Tanzania / Kenya

2000

2 1.0

1 1 1960

1980

2000

3 Sisal, Uganda

1960 1.5

1980

2000

Woods, non-coniferous

1960 1980 2000 Tropical Logs, UK import price, f.o.b 1.5

2 1.0

1.0

1 1960

1980

2000

1.5 Tropical Logs, Gabon

2.0

1960

1980

2000

Sawn Wood

2

1.5 1.0

1.0 1960

1980

2000

2

Plywood, cubic metre

1980

2000

1980

2000

3

1960

1980

2000

1980

2000

Rubber

2

1

1 1980

1980

2000

1980

2000

Leaf Tobacco

1.0

Cattle Hides

1960

1960 1.5

Linseed Oil

1 1960

2000

2

1 2

1980

Plywood, sheet

1 1960

3

1960

2000

1960

Appendix 3.11. Real Prices of 17 Agricultural Raw Materials

1960

3

2.0

Phosphate Rock

1.25

1960

1980

Tungsten Ore

0.75 1960

2000 1.5

Iron Ore

1.00

1.0

1

3

Manganese Ore

1.5

2

1980

2000

1960 3

Aluminium, high grade

1980

2000

Copper, Grade A

2

2 1.0

1

1 1960

1980

2000

1960 2.0

2.0

Copper, wire bars

1.5

1980

1960

2000

1980

2000

1980

2000

2.0

Nickel

Nickel cathodes

1.5

1.5

1.0

1.0

1.0 1960 3

1980

2000

2 1 1960 1.50

1960 2.5 2.0 1.5 1.0

Lead

1980

2000 1.0

1.25

2000

1960 3

Zinc, special high grade

2 1 1960

Zinc, Prime Western

1980

Refined Lead

1980

2000

Tin, LME

1960

1980

1.0

2000 Tin, ex-smelter price

1.00 0.5

0.75 1960 1.5

1980

2000

0.5 1960

1980

2000

1980

2000

3

Gold

Silver 2

1.0

61

1 0.5 1960

1980

2000

1960

Appendix 3.12. Real Prices of 17 Agricultural Raw Materials

1960

1980

2000

2.50

RCOF1 RCOF3

2.25

RCOF1 = Coffee, Columbian mild Arabicas RCOF2 = Coffee, Brazilian and other natural Arabicas RCOF3 = Coffee, Robustas RCOF4 = Coffee Composite Indicator (1976)

RCOF2 RCOF4

2.00 1.75 1.50 1.25 1.00 0.75 0.50 0.25 1960

1965

1970

1975

1980

1985

1990

1995

2000

Appendix 3.13. Real Price of Four Different Types of Coffee RWHAR

RWHAR = Real Price of Wheat (Argentina) RWHUS = Real Price of Wheat (US,red hard)

RWHUS

2.0 1.5 1.0

1960

1965

1970

1975

1980

1985

1990

1995

2000

RWHAR × RWHUS 2.0 1.5 1.0

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

2.0

Appendix 3.14. Real Price of Two Types of Wheat 2.50

RCRSO RGRNO

RSUNO

RCRSO = Real Price of Crude Soybean Oil RSUNO = Real Price of Sunflower Oil RGRNO = Real Price of Groundnut Oil

2.25 2.00 1.75 1.50 1.25 1.00 0.75 0.50 1960

1965

1970

1975

1980

1985

1990

1995

2000

Appendix 3.15. Real Price of Crude Soybean, Sunﬂower, and Groundnut Oil

Appendix 3.16. Data Set for 13 Commodity Prices (the Updated Grilli-Yang Series: 1900–2001) Year

Cocoa

Coffee

Tea

Bananas

Sugar

Rice

Wheat

Maize

Cotton

Jute

Palm oil

Copper

Tin

MUV

1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933

8.990 8.460 8.510 8.560 8.780 8.620 8.830 10.710 7.440 6.370 5.990 6.320 6.690 7.440 6.640 8.940 7.600 5.990 6.900 9.960 7.280 4.170 4.920 4.070 4.070 5.080 6.160 8.460 6.850 5.570 4.390 2.780 2.360 2.360

4.594 3.598 3.044 3.100 4.317 4.594 4.428 3.598 4.594 4.871 5.757 7.804 8.856 7.196 6.365 5.314 5.867 5.646 7.030 13.727 10.572 5.757 7.915 8.192 11.790 13.616 12.343 10.351 12.841 12.232 7.140 4.871 5.867 5.037

18.814 17.994 19.067 14.521 17.425 16.920 14.837 14.521 13.448 14.710 15.152 15.279 15.468 15.658 15.658 15.152 15.152 19.319 22.602 21.693 17.381 15.092 21.149 27.052 27.656 27.486 29.650 29.206 25.698 25.034 23.370 17.496 10.422 15.675

13.010 13.470 13.940 14.430 14.940 15.470 16.010 16.580 17.164 16.606 17.013 18.100 18.889 18.635 18.965 18.504 19.207 22.065 26.692 24.092 27.067 24.924 23.673 25.037 25.974 29.839 31.406 31.319 30.676 30.858 30.926 29.376 27.847 28.439

33.190 26.857 21.462 22.986 30.609 32.720 24.863 26.153 30.023 29.554 31.900 35.418 30.609 22.869 30.961 38.819 51.250 54.182 49.726 59.343 140.150 36.356 32.838 58.991 44.800 26.270 26.036 30.961 25.567 20.172 14.425 13.018 8.327 11.367

22.771 20.467 19.789 25.035 20.877 23.488 27.608 29.637 30.746 24.195 24.153 31.609 36.468 28.552 26.907 27.253 25.154 21.707 24.719 26.553 31.241 35.767 39.560 37.559 40.847 41.170 44.269 41.005 37.098 37.027 29.171 16.366 14.392 12.232

20.383 19.533 22.364 22.364 26.044 25.478 21.515 24.912 29.441 30.857 27.177 26.894 27.743 22.647 28.309 35.953 39.067 62.280 61.714 61.714 65.960 41.331 34.537 30.008 35.669 46.427 42.464 41.897 38.217 37.934 26.894 16.136 13.588 15.853

16.770 21.465 27.628 19.705 21.130 22.346 19.663 22.262 29.054 28.341 23.813 25.490 28.665 25.990 29.472 30.916 34.823 70.495 68.457 67.692 60.345 23.994 26.287 34.738 40.726 43.783 31.765 36.649 41.448 39.622 34.823 21.955 12.952 16.902

16.990 17.090 17.650 21.520 20.670 19.260 21.240 21.520 20.580 23.510 27.750 24.170 21.900 24.070 20.860 19.730 29.550 46.160 57.210 63.530 50.030 32.850 41.440 52.770 51.260 41.440 32.190 32.190 36.250 32.570 23.980 14.630 12.370 16.990

13.900 12.460 11.970 13.110 13.610 17.960 22.780 20.390 15.060 12.650 14.380 19.710 20.410 25.750 26.660 20.030 29.480 37.460 37.370 44.240 32.760 21.200 27.220 23.730 28.030 47.890 42.560 31.820 32.820 31.020 19.400 14.390 11.260 12.600

16.580 15.700 16.570 16.860 16.570 16.280 18.310 19.770 16.570 17.440 21.220 20.930 20.060 21.510 23.550 24.710 35.470 52.030 97.380 52.330 36.630 20.350 21.510 22.090 23.840 27.030 25.000 23.260 23.550 23.840 18.900 13.950 11.050 11.050

21.724 21.616 14.800 17.765 17.202 20.919 25.870 26.836 17.725 17.417 17.095 16.611 21.925 20.489 18.248 23.186 36.497 36.470 33.049 25.078 23.428 16.773 17.953 19.349 17.470 18.839 18.517 17.336 19.550 24.300 17.417 10.895 7.460 9.419

4.780 2.676 4.283 4.491 4.475 5.014 6.367 6.103 4.710 4.752 5.455 6.759 7.371 7.075 5.484 6.170 6.952 9.881 14.189 10.125 7.717 4.784 5.204 6.821 8.023 9.256 10.437 10.288 8.063 7.220 5.067 3.912 3.521 6.253

14.607 13.858 13.483 13.483 13.858 13.858 14.607 15.356 14.232 14.232 14.232 14.232 14.607 14.607 13.858 14.232 17.603 20.974 25.468 25.966 28.839 24.439 21.723 21.723 21.723 22.097 20.974 19.850 19.850 19.101 18.727 15.356 12.734 14.232 (Continued )

Appendix 3.16. (Continued ) Year

Cocoa

Coffee

Tea

Bananas

Sugar

Rice

Wheat

Maize

Cotton

Jute

Palm oil

Copper

Tin

MUV

1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967

2.780 2.680 3.640 4.500 2.780 2.570 2.730 4.070 4.760 4.760 4.760 4.760 6.160 18.680 21.250 11.560 17.180 19.000 18.950 19.860 30.940 20.070 14.610 16.380 23.710 19.590 15.200 12.100 11.240 13.540 12.530 9.260 13.060 15.580

6.144 4.926 5.258 6.089 4.262 4.096 3.930 6.255 7.417 7.417 7.417 7.528 10.240 14.779 15.000 18.103 25.862 30.517 29.483 29.483 39.828 31.034 35.690 35.172 25.862 22.241 21.207 19.655 18.621 18.103 24.310 23.793 21.724 20.172

21.194 19.996 20.592 23.764 22.264 19.355 19.770 25.300 32.973 29.655 28.964 28.964 27.443 38.020 40.439 43.072 38.370 48.151 40.157 48.057 69.405 65.831 63.668 58.401 60.470 60.000 60.659 58.025 58.684 55.674 56.521 55.110 53.605 54.075

28.188 28.410 27.423 26.327 27.103 28.385 30.986 32.080 33.169 34.676 37.051 38.652 43.921 45.998 46.981 52.050 54.383 54.383 55.126 55.126 56.647 55.903 56.647 59.621 55.126 49.177 48.434 46.947 44.716 56.674 57.391 53.639 52.152 53.639

13.956 18.530 20.289 20.641 17.005 17.709 15.950 19.820 29.671 28.616 28.968 34.480 41.634 56.293 49.608 48.788 58.404 66.496 48.905 39.992 38.233 37.998 40.813 60.515 41.047 34.831 36.825 34.128 34.949 99.686 68.847 24.863 21.814 23.338

15.912 20.785 20.405 21.633 19.850 19.208 23.088 25.814 28.353 29.628 29.628 29.628 34.177 54.719 57.992 50.082 42.235 44.614 48.290 53.975 48.785 43.718 42.297 42.389 43.965 40.844 38.527 42.173 47.209 44.274 42.544 42.111 50.422 63.587

21.515 24.063 26.611 37.934 28.309 17.552 18.684 18.684 20.949 34.254 36.519 39.350 64.545 75.019 67.376 56.028 49.645 57.447 60.283 54.610 47.518 45.886 45.886 44.539 43.759 45.319 44.539 45.106 46.879 47.447 49.787 46.666 49.787 49.220

27.561 34.526 35.502 43.741 23.144 21.233 24.461 29.939 35.375 43.868 48.115 49.559 69.306 87.524 40.738 60.771 65.586 69.342 60.289 57.978 56.148 46.998 49.695 45.843 45.843 44.398 41.702 44.206 49.503 52.681 53.740 52.970 57.207 48.058

21.900 22.660 22.940 20.200 16.340 17.560 19.540 27.090 35.020 36.630 37.950 42.480 54.470 62.020 58.910 56.920 67.680 75.330 66.930 60.890 60.420 60.510 58.620 57.110 57.870 56.070 53.620 56.260 59.190 58.720 55.980 52.490 43.800 36.250

14.180 16.530 17.460 19.430 17.270 23.540 22.360 21.230 19.940 26.620 32.410 31.600 39.560 64.330 77.920 56.710 62.800 93.500 56.020 44.480 52.240 47.330 49.750 59.510 52.010 51.290 74.580 91.500 62.070 61.050 66.860 71.200 79.230 56.560

15.700 22.380 22.670 25.000 19.700 20.350 20.930 28.200 34.590 25.000 25.000 25.000 25.000 51.740 63.370 46.800 42.440 67.150 40.120 35.470 36.340 37.790 43.600 44.190 41.860 42.440 41.280 42.440 40.410 40.700 41.280 45.930 43.900 43.020

11.311 11.607 12.707 17.672 13.418 14.706 15.162 15.833 15.806 15.806 15.806 15.806 18.544 28.124 29.573 25.763 28.500 32.472 32.472 38.644 39.838 50.304 56.114 39.690 34.565 41.837 43.005 40.147 41.059 41.059 42.884 46.990 48.533 51.297

8.344 8.061 7.425 8.688 6.763 8.045 7.967 8.317 8.314 8.314 8.314 8.314 8.720 12.463 15.868 15.883 15.275 20.318 19.261 15.323 14.683 15.147 16.214 15.390 15.209 16.316 16.218 18.116 18.330 18.650 25.197 28.491 26.232 24.531

16.854 16.479 16.479 16.854 17.603 16.105 17.603 18.727 21.723 24.345 27.715 28.464 28.839 34.831 35.581 33.333 30.337 35.955 36.704 35.206 34.457 34.831 36.330 36.704 36.330 36.330 37.079 37.453 37.453 37.453 38.202 38.951 39.700 39.700

1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001

18.410 24.460 18.310 14.340 17.290 34.470 52.510 40.040 58.500 108.390 97.410 94.200 74.400 59.410 49.600 60.570 70.110 65.660 60.980 58.760 44.657 35.497 36.355 34.351 31.489 32.061 40.077 40.936 41.794 46.374 48.092 33.779 26.050 30.630

20.172 20.690 26.897 23.276 25.862 32.069 34.138 33.621 73.965 124.660 85.862 89.483 80.690 66.207 72.419 68.048 74.550 75.530 100.520 58.660 72.258 56.996 46.980 44.595 33.625 37.202 78.935 79.412 64.150 99.444 71.066 52.942 45.787 32.671

44.577 41.473 46.740 44.953 44.859 45.141 59.812 59.060 65.549 114.640 93.386 91.975 95.173 86.050 82.447 99.339 147.520 84.420 82.290 72.879 67.634 76.324 76.702 69.523 75.569 70.279 69.145 61.966 62.722 77.836 77.458 68.767 71.035 60.455

51.408 53.639 55.903 47.690 54.383 55.734 62.224 83.382 86.965 92.846 97.104 110.050 128.100 135.640 126.510 144.930 125.000 128.378 129.050 123.310 161.486 184.797 182.770 189.189 159.797 149.662 148.311 150.338 158.784 169.932 166.216 144.932 143.243 196.959

23.221 39.523 43.979 53.010 87.137 112.940 351.360 240.420 135.810 95.229 91.477 113.290 336.240 198.900 99.053 99.534 61.310 48.090 73.030 81.563 118.280 150.538 150.538 107.527 107.527 118.280 145.161 155.914 139.785 134.409 107.527 75.269 96.774 102.151

62.286 57.744 44.490 39.856 45.448 108.140 167.460 112.180 78.630 84.099 113.540 102.360 134.060 149.230 90.532 85.596 77.930 66.740 65.540 71.705 91.238 98.485 89.262 96.509 88.274 77.404 88.274 105.731 111.660 99.802 100.132 83.004 66.535 56.983

47.163 45.603 44.822 45.319 50.567 104.400 148.010 128.580 105.740 82.127 95.602 122.270 135.320 139.290 118.040 120.220 117.310 104.350 88.238 86.703 111.256 129.671 104.350 98.980 115.860 107.420 115.092 135.809 159.595 121.998 96.678 86.703 87.470 97.445

47.287 51.910 56.244 56.244 53.933 94.382 127.130 115.180 108.250 91.782 96.983 111.240 120.670 125.970 105.280 130.970 130.870 108.050 85.710 74.099 102.885 107.692 104.808 102.885 100.000 98.077 103.846 118.269 159.615 112.500 98.077 87.500 85.577 86.538

35.680 35.400 37.850 48.050 49.273 84.825 88.567 72.351 105.408 96.676 97.923 105.408 127.861 115.387 99.794 115.387 111.645 82.330 66.114 102.913 87.320 104.160 113.516 104.784 79.835 79.835 109.774 132.851 110.397 109.150 89.815 75.469 81.083 66.114

61.850 69.160 66.340 72.750 69.430 67.170 83.720 83.140 70.640 89.550 104.740 105.710 98.370 88.520 82.120 86.140 157.370 173.060 79.080 94.500 100.749 101.656 111.186 99.502 76.016 73.860 80.509 99.570 123.690 82.233 70.546 75.063 75.924 89.743

42.150 36.340 46.220 43.900 34.880 48.260 77.910 75.580 59.300 88.370 101.450 110.170 99.130 95.640 73.550 87.210 126.900 87.210 43.520 58.012 73.487 58.857 48.767 57.007 66.256 63.566 88.790 105.607 89.295 91.817 112.838 75.842 52.131 48.095

56.154 63.776 77.422 69.009 67.922 78.992 102.850 85.258 92.343 88.304 87.901 123.790 137.370 112.380 97.791 104.440 89.580 87.980 86.670 112.460 167.439 183.269 171.300 150.515 146.782 123.102 148.456 188.932 147.683 146.525 106.435 98.970 116.667 101.544

Note: All price data are in nominal US dollars with 1977–79 prices. MUV is the prices index of manufactured goods exported by developed countries.

23.686 26.300 27.853 26.756 28.374 36.376 63.356 54.331 60.726 85.472 100.660 113.870 135.260 115.510 104.530 103.070 99.730 94.660 58.560 63.583 79.55 83.85 100.67 72.94 65.57 71.60 60.62 64.05 72.70 72.25 66.17 66.94 63.24 65.20

39.326 40.449 42.697 45.318 48.689 58.801 71.161 79.026 78.652 86.517 98.876 114.610 125.470 119.100 115.730 110.490 108.610 109.590 130.300 147.337 156.486 155.224 170.999 170.999 176.047 166.898 171.315 188.351 182.357 169.422 162.796 156.486 148.283 144.497

Appendix 3.17. UNCTAD Data on Commodity Prices Year

UVXM

Composite/ Aggregate

All Food

Food and Beverages

Food Only

Vegetable oils and oilseeds

Agri raw materials

Minerals and Metals

1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987

34.68208 35.4528 35.83815 35.83815 36.41618 37.28324 38.15029 38.15029 38.15029 38.15029 39.49904 42.00382 45.37572 53.46821 64.45087 73.12139 72.83237 80.0578 91.6185 104.0462 115.6069 109.5376 105.7803 102.3121 98.84393 100 119.3642 134.9711

45.20833 43.44167 43.25833 52.04167 51.525 48.36667 49.90833 47.66667 47.23333 51.8 53.575 51.63333 58.98333 95.375 138.9417 109.175 106.3833 117.3583 115.0833 132.7917 168.5167 140.4083 110.6417 117.95 113.1667 100.0083 104.0167 106.625

43.90833 43.10833 43.83333 58.55833 54.28333 45.79167 46.61667 47.475 45.83333 49.875 53.55833 53.075 64.2 107.25 169.0417 127.0417 115.1167 131.4333 125.0917 137.825 188.4083 154.3083 114.5083 122.525 117.6333 100.0083 107.4333 100.8167

30.475 28.05833 27.34167 28.36667 33.1 30.45833 31.05 30.43333 30.49167 32.36667 36.70833 32.26667 36.05 46.35833 54.48333 53.21667 96.125 169.0667 121.8917 125.6583 117.55 96.55 91.675 96.225 109.7167 100 124.2 80.70833

49.95 49.2 51.74167 75.25833 65.53333 51.30833 53.23333 55.71667 53.84167 59.61667 61.11667 62.03333 80.39167 138.8667 226.5167 168.15 130.15 119.3333 128.1417 144.0083 235.4833 189.6833 130.1583 137.7083 115.6083 100 109.475 115.8167

43.975 46.64167 41.75 43.99167 46.34167 52.61667 48.53333 44.90833 40.70833 41.2 53.81667 55.20833 47.75 87.825 141.9667 90.36667 84.375 107.9417 117.5083 134.625 116.825 110.2583 89.11667 106.8333 144.0917 100.0083 61.59167 72.6

54.26667 48.3 45.50833 45.46667 45.725 46.08333 46.1 42.73333 42.79167 45.84167 42.41667 42.2 46.78333 81.54167 84.90833 76.275 97.25 100.875 107.4583 124.8583 137.225 117.95 101.4667 108.3667 110.1417 100 103.5167 121.0917

41.85 40.70833 40.15833 40.19167 48.675 56.65 61.11667 51.75833 54.10833 61.075 61.75833 54.9 54.60833 75.21667 101.6167 87.60833 90.78333 93.50833 95.13333 125.725 140.6083 121.35 107.5083 113.2667 104.0167 100.0083 95.65833 110.8833

1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

143.3526 142.1965 156.6474 156.6474 161.2717 152.8902 156.9364 172.5434 167.052 155.2023 149.1329 143.3526 135.8382 132.3699 129.4798

134.8333 135.9583 128.8083 119.7917 115.85 110.3833 130.1667 143.3583 137.0667 136.3667 118.45 101.8583 104.4 101.0333 99.225

125.7333 126.8 117.3667 110.2667 106.575 107.9 129.85 135.9417 137.425 140.075 122.45 98.80833 97.31667 96.825 96.84167

81.725 69.99167 61.55833 56.85833 48.59167 52.31667 91.25 91.925 77.525 103.7 85.525 68.075 58.7 46.10833 50.125

151.8667 161.2667 151.6917 140.875 137.125 137.6833 151.9917 159.5333 169.625 162.475 139.5833 114.225 120.2833 126.3333 121.0917

96.075 84.925 74.04167 79.625 85.975 85.61667 107.4417 118.425 113.05 111.7083 120.3833 91.60833 71.06667 65.45 82.03333

129.825 132.4417 142.3083 133.6667 130.0167 120.8917 140.0917 161.3167 144.525 130.1083 116.3167 104.2 106.3833 104.1833 96.925

161.7417 161.9583 148.1667 133.95 129.175 109.0333 123.7333 149.1667 130.7417 131.4667 109.8083 107.9583 120.9917 109.4667 107.0667

4 Analysis of Movements in the Productivity and Prices of Selected Tropical Commodities in Developing Countries, 1970 to 2002 Euan Fleming, Prasada Rao, and Pauline Fleming

The international community accepts that a secular decline has occurred in the terms of trade of most commodities produced by developing countries, a trend that is conﬁrmed in this study. Yet it has been more or less indifferent to the fate of commodity producers in these countries, particularly since the collapse of international commodity agreements (such as the International Coffee Agreement and the International Cocoa Agreement given our special interest in agricultural commodities). Attempts have been made, albeit not very successfully, to reduce the degree of commodity price ﬂuctuations. But the problem of a secular decline in commodity prices has not been tackled in earnest.

4.1. Background to the Study 4.1.1. The setting of the research problem UNCTAD (2004b, p. 22) observed that the ‘net effect of the secular decline in prices depends on the extent to which world market prices are transmitted to producers and whether higher export volumes (for example, through productivity and yield improvements) make up for falling prices’. In this context, there is a need for evidence of the problem of declining prices faced by commodity exporting developing countries, of which there are many in the African, Caribbean, Paciﬁc, and Asian regions. A pertinent issue is the role of productivity in commodity production and its effects on individual countries. Two related matters concerning productivity

68

Analysis of Movements in Productivity and Prices trends are important. First, the European Commission recently communicated to the Council and the European Parliament that: Commodity prices demonstrate a long-term declining price trend. This trend has been driven mainly by signiﬁcant productivity gains, which enable producers to accept lower prices for their products. Other factors have also increased production: pressure on countries to earn more foreign exchange but few potential activities with which to do so; devaluation of national currencies of many commodity-producing countries following structural adjustment programmes; entry of new areas into production; and production subsidies in certain countries. The demand for commodities has not kept up with the increase in supply. (Commission of the European Communities 2004, p. 8)

If it is true that commodity export volumes have increased signiﬁcantly, the next step is to establish whether or not the above statements are true—that (a) ‘This trend has been driven mainly by signiﬁcant productivity gains’ and (b) producers are able ‘to accept lower prices for their products’. The validity of the ﬁrst assertion depends on the extent to which export expansion is due to increased input use on the farm or increased productivity in farm production. An empirical study of agricultural productivity growth is needed to conﬁrm this statement. For the second assertion to be valid, economic growth achieved through export expansion should not be immiserizing if productivity growth is high enough. It is of interest to learn about the pattern of productivity improvement across commodity-producing countries in the developing world. In particular, it is important to know whether increases in productivity have compensated for the decline in producer and export prices of commodities in Commonwealth countries that rely considerably on commodity exports to generate economic development. Much of the evidence presented in this report is used to test these two propositions. The greatest and most consistent concern about the ability of the developing world to achieve signiﬁcant productivity gains in agriculture has been focused on the African continent and, in particular, sub-Saharan Africa. The problems faced by agricultural producers in this region are well known and routinely spelt out. In a recent report to the United Nations on African agriculture, a panel of 18 experts employed by IAC, a Dutch non-government organization, observed that the sector was stagnant (M2 Presswire 2004, p. 1): The panel has learned that, among other things, Africa faced irregular rainfall and irrigation systems; low investment in agriculture; a wide variety of crops, and a lack of knowledge— largely due to brain drain, with some 50 per cent of the people qualiﬁed to make decisions and promote innovations towards alleviating food insecurity had left the country.

Many other factors could be added to this list such as declining soil fertility and poor to non-existent infrastructure in many areas, widespread disease and

69

The Issue of Declining Commodity Prices malnutrition in the rural population, and civil disturbances affecting commerce in the rural areas of many countries at different times.

4.1.2. Research objectives The research objectives are to investigate whether: . producers of tropical commodities in developing countries have compensated for falling producer prices by increasing total factor productivity . falling export prices have been compensated by rising total factor productivity of tropical commodities at the national level in developing countries.

4.1.3. Plan of the study 4.1.3.1. INDIVIDUAL COMMODITY STUDIES OR A SECTOR-WIDE ANALYSIS? The research objective entails two strands of estimation: changes in producer prices of selected tropical commodities and changes in total factor productivity in the production of tropical commodities. In this section, two approaches are considered to carry out the estimations required to achieve the second part of the research objective: . A series of individual studies of productivity change in commodity production across developing countries . Sector-wide analyses of productivity change by developing country. The former approach has the advantage of providing numerous speciﬁc measures of productivity change by commodity and by country. If enough of these studies were eventually done, a picture could be built up of productivity trends for individual farm enterprises across the developing world. But three major shortcomings make this approach infeasible. First, the data do not exist in enough countries for this approach to be followed for even one commodity let alone a set of commodities. Second, even where data do exist, the time series are unlikely to be long enough to enable a proper longitudinal study of productivity change to be undertaken. Finally, this partial approach ignores factors that inﬂuence productivity change on the farm as a whole, such as substitution of factors of production between enterprises. The latter approach of a sector-wide analysis of productivity change largely overcomes the problems outlined above. A sufﬁciently long data set is available (over three decades) across a wide range of developing countries, and the measures obtained will encapsulate the effects of resource use decisions among enterprises on total factor productivity. The main disadvantage is that productivity measures cannot be obtained for speciﬁc enterprises of interest. However, this can be overcome to a large extent by including as an explanatory

70

Analysis of Movements in Productivity and Prices variable the proportion of agricultural output contributed by the commodities of interest when regressing productivity change on a set of explanatory variables. This approach enables a measure of the rate of change in productivity to be made for a one per cent increase in the proportion of agricultural production devoted to each tropical commodity of interest. 4.1.3.2. TASKS UNDERTAKEN The following tasks were undertaken in the study: (a) Calculation of the growth in output and export of selected tropical commodities (coffee, cocoa, copra, palm kernel oil, coconut oil, palm oil, rice, cotton and sugar) in less developed countries from 1970 to 2002, with particular emphasis on Commonwealth and African countries. (b) Estimation of the rates of change in real export unit values of the selected commodities in (a) from 1970 to 2002, examining the movements in these prices and relating them to movements in corresponding world import prices. (c) Estimation of the rates of change in total factor productivity and labour productivity in agriculture from 1970 to 2002 in these less developed countries, with particular emphasis on Commonwealth and African countries producing and exporting the selected tropical commodities. (d) Comparison of the rates of change in productivity in the production of the selected tropical commodities with those for the whole agricultural sector. This is to be done by regressing change in total factor productivity and labour productivity separately on a set of explanatory variables that includes change in the ratio to output of each of the selected commodities in (a). (e) Comparison of the rates of change in productivity in the production of the selected tropical commodities, from (c), and rates of changes in their prices, from (b), for the period from 1970 to 2002. (f) Assessment of the single factoral terms of trade effects for the selected tropical commodities from (a) to test the proposition that the revenueenhancing effects of productivity growth have been outweighed by the revenue-reducing effects of declining commodity prices. (g) Review of the empirical literature on ‘immiserizing growth’ for evidence to test the proposition that output growth in African countries has led to welfare losses. (h) Identify those less developed countries with highest productivity gains in agriculture and compare them with countries achieving lowest (or negative) growth rates. (i) Identify those less developed countries that have managed to achieve agricultural productivity growth rates at least on a par with the rate of decline in real commodity prices.

71

The Issue of Declining Commodity Prices (j) Make policy recommendations where less developed countries have not been successful in realizing productivity gains and the secular decline in commodity prices is having a negative impact on the agricultural economy.

4.2. Review of Commodity Production and Export of Selected Tropical Commodities, 1970 to 2002 4.2.1. Importance of selected tropical commodities in the domestic economy Considerable differences exist in the importance of the selected commodities in the domestic economies of the countries under review. These differences are illustrated in Table 4.1 for countries in which the value of selected commodity exports is at least one per cent of the value of total exports. Proportions for the full list of countries are presented in Appendix 4.1. A number of African ˆ te d’Ivoire, Ghana, Rwanda and Uganda, and countries, notably Burundi, Co Central American countries have relied heavily on the commodities to contribute to both total export earnings and agricultural output. In other countries, such as Nigeria and Indonesia, the commodities have contributed little to export earnings but have been especially important to their large agricultural sectors. The commodities are of little importance for either export earnings or agriculture’s contribution to the economy in a number of countries, shown in Appendix 4.1. In no country is there the situation where the commodities contribute substantially to export earnings but not to agricultural output. We test the assertions made by the Commission of the European Communities (2004) by assembling evidence on the extent to which the quantities of tropical commodity exports have increased. The production and export of selected tropical commodities is reviewed over the period from 1970 to 2002. The focus of our study is a set of commodities of particular relevance to Commonwealth countries, especially to many in Africa. The commodities are coffee, cocoa, lauric oils (comprising copra, palm kernel oil and coconut oil), palm oil, rice, cotton and sugar. Table 4.2 contains a summary of trends in the export quantities of the selected tropical commodities in all countries, Commonwealth countries, African countries and African Commonwealth countries included in the study. Export quantity trends are described separately for indices of the tree crops (coffee, cocoa, palm oil, and lauric oils) and ﬁeld crops (rice, cotton and sugar). These trends are speciﬁed for a period of 33 years, from 1970 to 2002. Quantities are implicit volumes in that they are expressed in values normalized on 1990 international average prices. The indices are calculated using the Fisher index procedure.

72

Analysis of Movements in Productivity and Prices Table 4.1. Contributions by Selected Commodities to Export Earnings and Agricultural Output Export values of selected tropical commodities as a proportion of the total: Country Uganda Burundi Rwanda El Salvador Ghana ˆ te d’Ivoire Co Guatemala Honduras Nicaragua Colombia Costa Rica Kenya Cameroon Papua New Guinea Madagascar Dominican Republic Nepal Sierra Leone Central African Republic Congo, Republic of Togo Ecuador Haiti Malaysia Benin Brazil Peru Indonesia Malawi Bolivia Zimbabwe Guinea Jamaica Nigeria

FOB exports (%)

Agricultural output (%)

79.24 77.72 66.27 41.53 40.15 34.88 26.75 21.38 20.48 20.17 18.91 17.84 15.40 14.55 13.21 12.63 10.10 9.76 8.44 8.17 6.71 6.56 6.47 6.24 5.23 3.93 3.05 2.81 2.69 1.90 1.35 1.09 1.06 1.01

81.50 81.48 70.93 79.93 87.23 62.96 40.44 29.21 28.08 59.85 29.94 28.29 59.11 77.82 23.91 24.33 45.59 80.10 28.01 81.54 28.83 22.64 56.74 41.24 14.71 14.09 36.81 26.92 2.90 8.05 3.35 26.67 5.59 59.99

The output in Table 4.2 was generated with a series of ordinary least squares regression equations where the natural logarithm of the export quantity of interest was regressed on a trend variable. The estimated trend coefﬁcients are of particular interest. Each coefﬁcient reported for the trend can be interpreted as a percentage annual change. For example, in the ﬁrst section of Table 4.2 for all countries, the coefﬁcient for the trend variable in the second column under the total commodities heading is 0.018. This coefﬁcient means that the index of the export volume of all selected commodities increased on average by 1.8 per cent per annum between 1970 and 2002. The third to ﬁfth columns in Table 4.2 provide evidence that can be used to assess whether the change in export quantity is statistically signiﬁcant. Figures

73

The Issue of Declining Commodity Prices in the ﬁnal column showing the p (probability) values are the best ones on which to focus. A p value of 0.01, for instance, indicates that the relevant coefﬁcient in the second column is signiﬁcantly different from zero at the 1 per cent level (a high probability). Many p values for trend coefﬁcients are less than 0.001, indicating an extremely high probability that they differ from zero. Finally, the R2 values indicate the proportion of the variation in the export quantity that is explained by the regression model including the trend as an explanatory variable. Using the ﬁrst regression as an example again, the R2 value of 0.839 indicates that 83.9 per cent of the variation in the quantity index of all export commodities is explained by the trend variable. The trends in the export quantities of the selected commodities are discussed for all countries. The discussion is initially based on results for all selected commodities, followed by separate discussions for tree crops and ﬁeld crops. Similar discussions are then provided for particular sub-groups of countries.

4.2.2. Export quantities of the selected tropical commodities in all countries Movement in the export quantities index of the selected tropical commodities for all countries in the sample for the period from 1970 to 2002 are presented in Figure 4.1. The linear trend lines reported in Table 4.2 are also included for each series. Clear evidence is presented of signiﬁcantly increasing trends in both tree and ﬁeld crops over the study period. Tree crop exports display a higher rate of expansion that is to a large extent the outcome of substantial increases in tree plantings in response to a series of spikes in price about a decade apart. The planting response of individual producers was augmented by government encouragement to increase plantings as part of national planning efforts to bring about agricultural development, which tended to be intensiﬁed in periods of very high prices. This trend is evident in Figure 4.1 following mid-decade commodity booms. The average annual increase in the quantity index for all selected commodities is 1.8 per cent (Table 4.2). The average annual increase for tree crops of 2.7 per cent is around three times that for ﬁeld crops at 0.9 per cent. The rates for both sets of commodities and the rate for the index of all commodities are highly signiﬁcant in statistical terms.

4.2.3. Export quantities of the selected tropical commodities in sub-groups of countries Export quantity indices of the selected tropical commodities for Commonwealth countries in the sample are presented in Figure 4.2. The annual rate of increase in the index for all commodities, reported as 3.3 per cent (Table 4.2), is substantially higher than that for non-Commonwealth countries. The annual

74

Analysis of Movements in Productivity and Prices Table 4.2. Estimates of Trends in Export Quantities of Selected Commodities, 1970 to 2002 Variable

Trend coefﬁcient

p-value

All countries: Total commodities Tree crops Field crops

0.018y 0.027y 0.009y

<0.001 <0.001 0.001

Commonwealth countries: Total commodities Tree crops Field crops

0.033y 0.037y 0.025y

<0.001 <0.001 <0.001

African countries: Total commodities Tree crops Field crops

0.010y 0.007y 0.015y

<0.001 <0.001 <0.001

0.002 0.001 0.007*

0.369 0.555 0.039

African Commonwealth countries: Total commodities Tree crops Field crops y

Signiﬁcant at 1 per cent level *Signiﬁcant at 5 per cent level

Export Quantity - All Countries 1.8 1.6

Index (1990=1.00)

1.4

Total quantity Tree crop quantity Field crop quantity Linear (Field crop quantity) Linear (Total quantity) Linear (Tree crop quantity)

1.2 1.0 0.8 0.6 0.4 0.2

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

0.0

Figure 4.1. Export Quantity Index for Selected Commodities in All Sampled Countries, 1970–2002

rates of increase are higher for both tree crops and ﬁeld crops, at 3.7 per cent and 2.5 per cent, respectively. As for the rates of increase for all countries, both rates are highly signiﬁcant. Yet there is a substantial range of export growth rates. While Commonwealth countries as a group experienced rapid growth in

75

The Issue of Declining Commodity Prices Export Quantity - Commonwealth Countries 2.5

Index (1990=1.0)

2.0

Total quantity Tree crop quantity Field crop quantity Linear (Field crop quantity) Linear (Total quantity) Linear (Tree crop quantity)

1.5

1.0

0.5

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

0.0

Figure 4.2. Export Quantity Index for Selected Commodities in Sampled Commonwealth Countries, 1970–2002

Export Quantity - African Countries 1.6 1.4

Index (1990=1.0)

1.2

Total quantity Tree crop quantity Field crop quantity Linear (Field crop quantity) Linear (Tree crop quantity) Linear (Total quantity)

1.0 0.8 0.6 0.4 0.2

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

0.0

Figure 4.3. Export Quantity Index for Selected Commodities in Sampled African Countries, 1970–2002

tree crop exports, African countries realized only slow growth and African Commonwealth countries achieved no signiﬁcant growth. Export quantities of all selected tropical commodities, selected tree crops and selected ﬁeld crops for African countries in the sample are presented in

76

Analysis of Movements in Productivity and Prices Export Quantity - African Commonwealth Countries 1.6 1.4

Index (1990=1.0)

1.2 1.0 0.8 0.6 0.4 0.2

Total quantity Tree crop quantity Field crop quantity Linear (Field crop quantity) Linear (Tree crop quantity) Linear (Total quantity)

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

0.0

Figure 4.4. Export Quantity Index for Selected Commodities in Sampled African Commonwealth Countries, 1970–2002

Figure 4.3. The overall rate of increase in export volume of 1.03 per cent is considerably lower than that for all countries and Commonwealth countries, because of the much lower annual growth rate for tree crops of a mere 0.70 per cent. This rate is little more than one-quarter the annual rate of export growth in tree crop exports for all countries in the sample. The annual growth rate for ﬁeld crops of 1.50 per cent is quite a bit higher than the rate for all countries. The picture painted above for export quantities in African countries is bleaker for African Commonwealth countries. Exports by these countries of all selected tropical commodities and of tree crops in Figure 4.4 show no signiﬁcant growth over the study period. The overall rate of increase in export volume of ﬁeld crops is signiﬁcant but low, at 0.66 per cent per annum. While similar patterns of growth in export quantity indices have taken place between sub-groups of countries, quite a few countries, particularly in Africa, have not been beneﬁting from volume increases as much as others. The degree of volatility was higher in the ﬁnal decade of the study period, most likely due to the effects of structural adjustment.

4.3. Is there Evidence of Immiserizing Growth in Developing Country Agriculture? 4.3.1. The mechanisms of immiserizing growth The so-called ‘immiserizing growth’ hypothesis was originally expounded by Bhagwati (1958) who developed a simple two-country, two-good trade model 77

The Issue of Declining Commodity Prices to explain how immiserizing growth can occur. Kuhnen (1987) described the process as follows: This theory follows the argumentation of the theory of circular deterioration of terms of trade and concludes that countries, in order to improve their balance of trade, have to increase their exports to compensate for falling prices. This means a further deterioration of terms of trade. The unchanged structure of supply intensiﬁes the structural dependency and, regardless of growth, there is no development but only ‘immiserizing growth.’ This situation is especially pertinent for countries with agrarian monoculture. As a consequence, Bhagwati later asked for a speedy industrialization including heavy industry for larger countries.

The terms of trade must move against a country for immiserizing growth to occur. It has often been argued that weakness in commodity prices has been the result of productivity gains in primary production. For example, it was mentioned above that the Commission of the European Communities (2004, p. 8) had asserted that commodity prices have been inexorably declining mainly as a result of substantial increases in agricultural productivity. Sawada (2003, p. 2) set immiserizing growth in the context of this study as follows: There is a well-known phenomenon in international trade theory where increasing welfare and positive economic growth do not coincide. This is the case of immiserizing growth. The prototypical example of immiserizing growth is where export-biased growth by poor countries worsens their terms of trade so much that they are worse off than if they had not grown at all.

He characterizes immiserizing growth as a theoretical extreme of the wellknown Prebisch-Singer hypothesis, which Prebisch (1950) and Singer (1950) set in terms of developing countries on the periphery and developed, industrialized countries in the centre. Cypher and Dietz (2004, p. 166) summed up the hypothesis as ‘the center nations gain doubly from new technology and trade with the periphery, while the periphery becomes worse off as a result of a deterioration in their terms of trade that results from the price movements on center exports and periphery exports’. Prebisch and Singer contended that two factors were reinforcing these technological change differentials and domestic and international market structures, namely substantially higher income elasticities of demand for manufactured products than for primary products and a low propensity to import in the USA (Cypher and Dietz 2004, pp. 166–167). In respect of the former, one could add that substantially higher income elasticities of demand also tend to be associated with services involved in the marketing of commodities. The validity of the Prebisch–Singer hypothesis has been the subject of controversy over many decades. This controversy continues today, but the evidence is accumulating that the terms of trade for commodities produced by developing

78

Analysis of Movements in Productivity and Prices countries are deteriorating. Cypher and Dietz (2004, pp. 168–9) collated recent empirical evidence based on trends in real commodity prices from two sources not known for their support of the Prebisch-Singer hypothesis, the World Bank and the International Monetary Fund. First, they used data compiled by the World Bank (2002) to show markedly declining real prices of commodities (excluding petroleum products) between 1980 and 2001 for developing countries, as a whole, and sub-Saharan African countries in particular. The price index fell from 100 in 1980 to a little above 50 for the former and just above 40 for the latter in 2001 (Cypher and Dietz 2004, p. 168). Second, the IMF (1994) reported increased magnitudes of the average annual decline in the terms of trade for raw materials: 0.78 per cent between 1957 and 1987; 1.52 per cent between 1968 and 1987; and, for 33 commodities, between 3.6 per cent and 4.2 per cent from 1979 to 1993 (Cypher and Dietz 2004, p. 169). Coffee provides a stark example of these factors at play in bringing about deteriorating terms of trade in commodity markets in that virtually all coffee beans are produced in the developing world whereas most of the valueadding activities and consumption take place in the developed world. Recent trends in the world coffee market reveal a dramatic decline in the share of the consumer price of coffee claimed by producing countries. McCorriston, Sexton and Sheldon (2005, p. 1) reported that, in ‘key export markets such as Europe and the United States, global coffee buyers, roasters and retailers . . . account for almost 60 percent of the share of ﬁnal sales value of coffee’. Choraria (2005) presents evidence of compression of the value chain for coffee and cocoa between the farm-gate and retail levels (that is, a decreasing share of the retail value going to the farmer), with the compression occurring overwhelmingly between the point of export and the retail level.

4.3.2. Evidence of immiserizing growth Empirical analysis of the presence of immiserizing growth in a developing economy is a complex undertaking that is beyond what is achievable in this study. Instead, we review the applicability of this theory and the empirical evidence of immiserizing growth in African countries that is available from other studies. Despite the accumulating evidence of a decline in the real prices of non-oil commodities, limited evidence has been marshalled to support the existence of immiserizing growth. Empirical results from the analysis by Sawada (2003), using macroeconomic data on consumption to reﬂect revealed preferences, yielded many instances of immiserizing growth in Commonwealth commodity-exporting countries of interest in this study during the period from 1970 to 1988. They include Barbados, Fiji, Ghana, Swaziland, and Tanzania during 1975–1980, and Papua New Guinea, Tanzania and Trinidad and Tobago during 1985–1988. Among African countries in the sample, one-half experienced

79

The Issue of Declining Commodity Prices increases in welfare during 1975–1980 and only 39 per cent experienced welfare increases in 1980–1985. Unfortunately, the data needed to analyse the welfare impact of economic growth in agricultural industries using revealed preferences in consumption are unavailable in the necessary disaggregated form. A suitable alternative method that can be used is to examine trends in the single factoral terms of trade, outlined below.

4.4. Research Agenda The research agenda comprises three main analytical components of analysis of change in developing country agriculture in a study period covering a little over three decades from 1970: . Estimation of trends in productivity . Calculation of changes in selected commodity prices at the producer, national and global levels . A comparison of productivity changes and price changes for commodities commonly produced in developing countries, and an assessment of the net effect of these changes on producer incomes in these countries.

4.4.1. Estimation of productivity change in developing country agriculture In this section, we begin by deﬁning two common measures of productivity: the comprehensive measure of total factor productivity (TFP) and the most important partial productivity measure for our purposes, labour productivity. Our focus is on the TFP growth rate rather than the rate of growth in labour productivity because it encompasses the use of all inputs. But labour productivity provides a useful benchmark against which to compare TFP indices. Also, a comparison of the two measures produces some interesting contrasts that highlight variations among developing countries in the processes of structural change on which they have embarked. After deﬁning the productivity concepts, we outline the major processes generating changes in agricultural productivity in developing countries. 4.4.1.1. DEFINING TOTAL FACTOR PRODUCTIVITY AND LABOUR PRODUCTIVITY TFP in an agricultural sector is the ratio of all outputs to all inputs used to produce that output. Not all agricultural inputs are included in analyses in this study. Nevertheless, they are sufﬁciently comprehensive to capture overall changes in input usage. TFP change is an amalgam of technological change, change in technical efﬁciency and change in scale efﬁciency (that is, efﬁciency

80

Analysis of Movements in Productivity and Prices Wool output

Production frontier after technological progress

Production frontier before technological change Average production function

0

Inputs

Figure 4.5. Production Functions, Technological Change, and Technical Efﬁciency Change

gains from changing the scale of production operations). Figure 4.5 demonstrates the difference between technological change and technical efﬁciency change, the components of TFP of particular interest in this study. Each dot represents the position of a farm in respect of the production frontier. If farms below the frontier were to move closer to it an increase would occur in technical efﬁciency. The average production function, represented by the dotted line, would become closer to the frontier. If the production frontier were to shift, technological change would have taken place. Technological progress is shown by a move upwards of the production frontier in Figure 4.5. Labour productivity in the agricultural sector is the output obtained per unit of labour input. The labour input used is number of people economically active in agriculture. It is not sufﬁciently speciﬁc to capture short-term changes in labour use, but is suitable for this sort of study in which a longterm view is taken of trends in productivity. 4.4.1.2. PROCESSES GENERATING PRODUCTIVITY CHANGE IN AGRICULTURE Because exports from developing countries are dominated by crop products, the focus of the analysis is on productivity in crop production. The main processes bringing about changes in TFP in tropical cropping systems are: . Improved production technologies, including improved crop varieties . Improved quality of agricultural output . Improvements in the technical efﬁciency with which agricultural production is undertaken

81

The Issue of Declining Commodity Prices . Changes in the scale of production operations on the farm . Changes in environmental factors, such as infrastructure, soil structure and fertility, pesticide residues, desertiﬁcation and salinity.

4.4.2. Calculating trends in prices for tropical commodities We reported above the widespread contention that a secular decline has been taking place in real commodity prices over the past few decades and calculated changes in export quantities consistent with the view that increased export volumes have a depressing effect on export prices. In this section, we outline the method used to measure whether such a decline in export prices has occurred and, if so, the extent of this decline. Trends in real prices of the selected commodities for analysis are calculated for export prices, measured as export unit values, and producer prices from 1970 to 2002. A comparison is made between the two price series to detect whether any divergence has occurred in trends.

4.5. Methods of Analysis 4.5.1. A model for estimating change in total factor productivity 4.5.1.1. REASONS FOR CHOOSING AN OUTPUT ORIENTATION Most of the inputs used in developing country agriculture are ﬁxed in the short run, offering limited opportunity to producers to alter their resource mix in production. The best examples of these types of inputs are land, operator and family labour, irrigation equipment and most other plant and machinery items. To this list we should add those tree crops that are of special interest in this study: coffee trees, cocoa trees, coconut palms, oil palms and, to some extent, sugar crops that tend to be ratooned for a few years. For this reason, an output orientation is preferred for applying data envelopment analysis that allows us to calculate TFP indices. 4.5.1.2. USE OF MALMQUIST INDICES TO MEASURE CHANGES IN TOTAL FACTOR PRODUCTIVITY We calculate Malmquist indices to measure changes in TFP in the agricultural sector in each selected developing country over the study period. In this respect, we follow the same methodological path for calculating agricultural TFP change in African countries as, ﬁrst, Coelli and Rao (2004) and, later, Nkamleu (2004). One of the major advantages of estimating a Malmquist index is that it assumes an underlying translog production function that allows ﬂexibility in the relations between outputs and inputs in production technology.

82

Analysis of Movements in Productivity and Prices As in our study, Coelli and Rao included more than African countries, covering developed and transitional countries, whereas Nkamleu conﬁned his study to African countries. We differ from Coelli and Rao by excluding developed and transitional countries from our data set. Our study period extends one year beyond the period chosen by Coelli and Rao and Nkamleu. Two models are estimated, one covering crop production in all developing countries in the sample and a separate model for the majority of countries for which livestock data are also available, which provide the preferred TFP estimates.

4.5.2. Explaining changes in total factor productivity in developing agriculture 4.5.2.1. MODEL ESTIMATION Two pooled econometric models were estimated with TFP change and labour productivity change as the dependent variables. The models were estimated for 57 countries. Data availability and production of the selected commodities guided our choice of countries to include in the data set. It transpired that 26 of the 83 countries for which we report productivity estimates were excluded from the set owing to either missing data or failure of a country consistently to produce any of the commodities of interest. The TFP model was estimated for the 32-year period from 1970 to 2001 using annual data. A lack of data for most explanatory variables meant that 2002 was omitted from the analysis even though productivity estimates were available. The pooling option in the SHAZAM econometric package was used, and the full cross-sectionally correlated and time-wise autoregressive option was applied (Lee, White, and Granger 1993). A different value of rho was estimated for each cross-section. Annual data were pooled from the 57 countries, giving a total of 1824 observations. The economic model for TFP is: TFP ¼ f1 (PIDC, GDPPC85, CONV, RURAL, ILLIT, RAIN, SOIL, PROPF1, PROPF2, PROPF3, PROPF4, PROPF5, PROPF6, PROPF7, DRL, DRA),

(1)

where: TFP is change in the total factor productivity; PIDC is the deﬂated commodity price index; GDPPC85 is gross domestic product per head in constant 1985 prices; CONV is the labour productivity; RURAL is the proportion of the population residing in rural areas; ILLIT is the illiteracy rate;

83

The Issue of Declining Commodity Prices RAIN is the long-term average annual rainfall; SOIL is the mean soil fertility for the country; PROPF1 is the proportion of crop production contributed by cocoa output; PROPF2 is the proportion of crop production contributed by coconut output; PROPF3 is the proportion of crop production contributed by the output of green coffee; PROPF4 is the proportion of crop production contributed by the output of oil palm fruit; PROPF5 is the proportion of crop production contributed by the output of paddy rice; PROPF6 is the proportion of crop production contributed by the output of cotton; PROPF7 is the proportion of crop production contributed by the sugarcane output; DRL is a regional dummy variable for Latin American and Caribbean countries; and DRA is a regional dummy variable for African countries. The model was estimated in double-logarithmic form so that the coefﬁcients can be interpreted as elasticities. The base region is Asia. The economic model for labour productivity is based on a Solow growth model. It takes the form: LP ¼ f2 (FERTPU, TRACTPU, LANDPU, PIDC, GDPPC85, CONV, RURAL, ILLIT, PROPF1, PROPF2, PROPF3, PROPF4, PROPF5, PROPF6, PROPF7, DRL, DRS, DRA, RAIN),

(2)

where: FERTPU is fertilizer per unit of labour; TRACTPU is tractors per unit of labour; LANDPU is agricultural land per unit of labour; DRS is a regional dummy variable for South Asian countries; and the other variables are as speciﬁed above. The model was estimated in double-logarithmic form so that the coefﬁcients can be interpreted as elasticities. The base region is South-East Asia. 4.5.2.2. PROPORTION OF SELECTED TROPICAL COMMODITIES TO TOTAL EXPORTS There are two main reasons for estimating these models. The ﬁrst reason is to determine whether productivity growth estimates diverge from the average in the production processes for any of the commodities of interest in this study. It

84

Analysis of Movements in Productivity and Prices is unlikely that productivity growth has been the same in the production of all selected commodities. We suspect that the productivity growth of tree crops has lagged behind other crops, for four main reasons. First, the genetic and other research work carried out on the production of these crops has tended to make less progress than that made in other parts of agriculture. Second, many tree crops have been planted on previously forested land. High yields from initially fertile soils in the 1960s and 1970s have not been sustained as soils in many areas have deteriorated with continuous production over a number of decades in humid tropical climates (Gallup and Sachs, 2000). Third, many tree plantations are getting older and trees have not been replaced, leading to declining yields. Finally, a number of countries have experienced increased incidences of pests and diseases affecting the yields of tree crops, especially cocoa producers. On the other hand, considerable technological advances have been made in the production and protection of rice and cotton crops. Rice producers have beneﬁted from Green Revolution research outputs and the widespread diffusion of improved and sustainable production methods in the post-Green Revolution era, while cotton producers have beneﬁted from technology spillovers from developed agriculture, in countries such as USA and Australia. Sugar production, on the other hand, has suffered from growing soil infertility in many major sugar-producing countries. In summary, we expect rice and cotton production to record above-average productivity gains and tree crops and sugarcane production to record below-average productivity gains. Second, the relationship between productivity growth and commodity price change is of interest. We undertake signiﬁcance tests of the proposition that a negative relationship exists. That is, given the expectation of commodity price decline over the study period, we test the hypothesis that TFP and labour productivity growth rates increase in response to a decline in the price index of the selected commodities as producers attempt to offset the revenue-reducing effects of price declines. Our expectation is that producers have little ‘room to manoeuvre’ by extracting productivity gains in the face of declining prices. 4.5.2.3. OTHER FACTORS A number of other variables are included in the estimated model. We expect a positive sign on the variable measuring gross domestic product per head. Our reasoning here is that more economically advanced countries are better able to provide infrastructural and institutional support, have more developed marketing systems, and have greater industrial capacity that can generate improved input use in agriculture. The 1970 labour productivity variable is included as a proxy variable for convergence. Labour productivity in 1970 is used rather than 1970 TFP because, while TFP growth is usually a robust measure, TFP at levels tends to be

85

The Issue of Declining Commodity Prices less reliable as a measure of productivity than is labour productivity. The conventional wisdom from a neoclassical economic standpoint underlying the inclusion of this variable is that those countries starting from a lower productivity base in their agricultural sectors at the beginning of the study period are expected to have more scope to ‘catch up’ to other countries. That is, the lower the labour productivity at the start of the period, the higher is the expected rate of growth in TFP. This view is not universally accepted. Gallup and Sachs (2000), for example, reported divergence rather than convergence among agricultural sectors in different ecozones from 1961 to 1998. In studies such as Alauddin, Headey, and Rao (2005), the proportion of the population living in rural areas has been argued to have a negative impact on TFP. A high proportion of rural population is associated with lack of economic development and the slow growth of alternative employment opportunities outside agriculture. This situation results in considerable under-employment of one of the major agricultural factors of production, labour, that continues to grow in many less developed countries faced with high population growth rates. An alternative argument could be put that agriculture plays a more prominent role in the general economy in these countries, and governments are expected to have concentrated their development efforts where a majority of the population live in the knowledge that overall economic development would be greatly handicapped without signiﬁcant productivity gains in agriculture. This strategic focus was lacking prior to the commencement of the study period when many governments were placing their faith and efforts in industrialization to the neglect of the farming community. It is suspected that this ‘blind spot’ to the need to create the conditions for rapid agricultural growth has not yet disappeared, or even lessened in some less developed countries. The illiteracy rate is an obvious candidate for inﬂuencing TFP. We postulate that the higher the illiteracy rate the lower the rate of productivity growth. Nkamleu (2004) found a signiﬁcant negative relationship in his analysis of a sample of African countries. Physical conditions such as soil quality and rainfall can inﬂuence the scope for productivity gains. The long-term average rainfall is included in the estimated model, particularly to pick up the effects on TFP of differences between dry and humid tropical zones. Alauddin et al. (2005) found that soil fertility was associated with higher rates of productivity growth, as expected. In this study, however, the coefﬁcient on the soil fertility index, which is the same variable as that used by Alauddin et al., was not signiﬁcantly different from zero and was excluded from the ﬁnal estimated model. Finally, three regional dummy variables were initially included in both the TFP and labour productivity models, for Latin America and the Caribbean, South Asia and Africa. Preliminary analyses for the TFP model indicated no signiﬁcant difference in TFP growth between the base, South-East Asia, and South Asia, so

86

Analysis of Movements in Productivity and Prices this dummy variable was excluded from the estimated model and Asia as a whole served as the base region. Its coefﬁcient was signiﬁcant, and retained, in the labour productivity model where South-East Asia is the base.

4.5.3. Relative effects of productivity change and price change 4.5.3.1. MEASURES TO COMPARE THE RELATIVE EFFECTS OF PRODUCTIVITY AND PRICE CHANGE Because comprehensive data do not exist on input costs or enterprise mix at the individual farm level in the countries under study, a deﬁnitive statement cannot be made on the existence or otherwise of a decline in real farm incomes resulting from long-term declines in commodity prices. But inferences can be drawn on the potential immiserizing impacts of commodity price declines at the agricultural sector level under a set of assumptions about agricultural production costs and enterprise mix. The central proposition to test in this study is whether productivity change in commodity production in the selected developing countries has been high enough to offset declines in commodity prices received by producers. Four terms of trade concepts are discussed in international trade text books that could be applied as measures for this purpose, but two are too crude to provide any meaningful comparison and one, while theoretically superior, is impractical because of its stringent data requirements. The crudest and simplest measure, the net barter (or commodity) terms of trade, takes no account of changes in input and output relations and is unsuitable. The income terms of trade take into account changes in output quantities but ignore input quantities. The single factoral terms of trade is a useful measure that is discussed in detail below. The ideal measure, the utility terms of trade, requires a measure of the utility gained from trade by producers of the good being traded. Measuring utility is out of the question, but it is possible to use revealed preferences so long as consumption data for producers are available. Sadly, as indicated above, this is not so. The single factoral terms of trade is considered suitable for the purpose of this study in that it incorporates changes in inputs and outputs, through changes in TFP, and changes in the net barter terms of trade. Estimates of productivity, output price and import price changes are needed to calculate the single factoral terms of trade in each country. 4.5.3.2. CALCULATING THE SINGLE FACTORAL TERMS OF TRADE The single factoral terms of trade is a measure of the returns to the factors engaged in the production of a commodity or group of commodities. An increase in this measure means that ‘more [goods and services] can be purchased for a given amount of employment time of the factors of production’ (Appleyard and Field 1998, p. 120).

87

The Issue of Declining Commodity Prices We deﬁne the single factoral terms of trade using the deﬁnition and notation of Perkins, Radelet, Snodgrass, Gillis and Roemer (2001, pp. 637–641). Perkins et al. (2001, p. 637) begin by deﬁning the net barter terms of trade (Tn ) in the usual manner as the ratio of an index of export prices (Pe ) and an index of import prices (Pm ). They then use the net barter terms of trade to deﬁne the single factoral terms of trade as TS ¼ (Pe =Pm )Ze , where Ze is total factor productivity. This equation simpliﬁes to TS ¼ Tn Ze . The single factoral terms of trade, TS , measures factor income relative to factor inputs and import prices, or TS ¼ (Pe =Pm )Ze , ¼ Tn Ze. Note that a rise in either the income or single factoral terms of trade implies an improvement in income or welfare relative to a country’s previous situation. But if, as often is the case, [the] index rises less than export volume, this also implies that exporting countries are sharing part of the potential gains with importing countries, as Prebisch and Singer suggest. (Perkins et al. 2001, p. 637)

Perkins et al. used a single product (copper in Zambia) to demonstrate an application of the single factoral terms of trade but observed that, although the index is intuitively appealing, it is rarely used because of a lack of data on productivity. As indicated above, we plan to overcome that deﬁciency in this study by estimating trends in agricultural productivity for the selected tropical commodities to calculate their single factoral terms of trade. This measure enables us to determine whether the decline in the prices producers receive for their exports is less than the percentage rise in productivity in their production, in which case, given the deﬁnition by Perkins et al. above, returns to the factors engaged in their production would increase. If the price decline is greater, returns to the factors engaged in their production would fall. The single factoral terms of trade as deﬁned above covers all activities in producing a product or group of products to the point of export. As our purpose lies in examining welfare implications for producers as a speciﬁc group, we deﬁne the single factoral terms of trade in terms of output prices (Pf ) rather than export prices and farm-level productivity (Zf ). The appropriate index to use as the denominator is the consumer price index (Pc ), which is the equivalent ‘import price index’ for participants in the agricultural economy as the import price index is for participants in the national economy. Our measure of the single factoral terms of trade is therefore deﬁned as TS ¼ (Pf =Pc )Zf .

4.6. Data and Variables The source of all production, producer price and export unit value data is the FAOSTAT-Agriculture website, supported by the Statistics Division of the Food and Agriculture Organization of the United Nations (FAO) (www.fao.org/faostat). The proportions of the output mix contributed by the selected tropical commodities

88

Analysis of Movements in Productivity and Prices were also derived using the FAOSTAT-Agriculture data. Other data were obtained from the various sources listed by Alauddin et al. (2005, pp. 48–9).

4.6.1. Period of analysis and country coverage 4.6.1.1. STUDY PERIOD In developing countries, the production cycle is very long for numerous agricultural commodities including some of special interest in this study (coffee, cocoa, palm oil and lauric oils). Therefore, it is desirable to make the study period as long as possible to allow for changes made in agricultural technologies to take effect in these industries. The study period chosen is from 1970 to 2002. 4.6.1.2. COUNTRIES TO BE STUDIED The countries included in the analysis were restricted to the developing world to place some limits on the heterogeneity of agricultural sectors. An effort was made to include as many Commonwealth and African countries as possible given the focus of the study. Of the 88 countries that were initially selected for inclusion on the basis of FAO data to estimate productivity, two countries (Ethiopia and Iraq) did not have a full set data, one country (Re´union) is part of a developed country (France) and two countries (Libya and Myanmar) have data sets that are not considered reliable. All were omitted from the productivity analysis. We have doubts about the input data for one country in particular (Burundi), but nevertheless included it because the labour productivity index provides a realistic alternative to the TFP index. Three countries included in the analysis are marginal to the interests of the study. First, South Africa and Chile have little in the way of tropical agriculture and a signiﬁcant component of each agricultural sector shares attributes with developed agricultural sectors. They are nevertheless included because of the contrast that we expect they demonstrate in the ability of their farms to generate substantial TFP gains and because they are located on continents otherwise comprising developing countries in which tropical agriculture predominates. It allows us to undertake a test that extends the proposition put forward by Gallup and Sachs (2000, p. 731), that agricultural productivity is lower in the tropics, to propose that productivity growth in the temperate agriculture is higher than in tropical agriculture. Turkey is also on the margin of the developed world and has predominantly temperate farming systems. But it also has a large peasant-dominated agricultural sector that has beneﬁted little from technological advances in agricultural production. The 83 countries included in the analysis are listed in Table 4.3 along with the selected commodities for which they have substantial exports. The commodities

89

The Issue of Declining Commodity Prices Table 4.3. Major Exports of Selected Commodities Country

Selected commodities

Algeria Angola Argentina Bangladesh Barbados Benin Bolivia Botswana Brazil Burkina Faso Burundi Cambodia Cameroon Central African Republic Chad Chile China Colombia Congo, Democratic Republic of Costa Rica ˆ te d’Ivoire Co Cuba Dominican Republic Ecuador Egypt El Salvador Fiji Islands Gabon Gambia, The Ghana Guatemala Guinea Guinea-Bissau Haiti Honduras India Indonesia Iran Jamaica Kenya Laos Lesotho Liberia Madagascar Malawi Malaysia Mali Mauritania Mauritius Mexico Mongolia Morocco Mozambique Namibia Nepal Nicaragua

None Coffee; cotton Cotton; rice; sugar Cotton Sugar Cocoa; cotton; palm oil Coffee; copra; palm oil None Cocoa; coffee; cotton; palm oil; palm kernel oil; sugar Cotton Coffee; cotton None Cocoa; coffee; cotton; palm oil Coffee; cotton Cotton None Coffee; rice; sugar; palm oil Coffee; cotton; palm oil; palm kernel oil; sugar Cocoa; coffee; palm oil; palm kernel oil Cocoa; coffee; palm oil; rice; sugar Cocoa; coffee; cotton; coconut oil; palm oil; palm kernel oil; sugar Coffee; sugar Cocoa; coffee; sugar Cocoa; coffee; palm oil; rice; sugar Cotton; rice Coffee; cotton; sugar Coconut oil; sugar Cocoa; coffee; palm oil None Cocoa; coffee; palm oil Coffee; cotton; palm oil; sugar Cocoa; coffee; cotton None Cocoa; coffee; sugar Coffee; cotton; palm oil; sugar Coffee; cotton; rice; sugar Cocoa; coffee; coconut oil; palm oil; palm kernel oil; rice None Cocoa; coffee; sugar Coffee; sugar Coffee None Cocoa; coffee; palm oil Cocoa; coffee; rice; sugar Coffee; cotton; sugar Cocoa; copra; coconut oil; palm oil; palm kernel oil; sugar Cotton None Sugar Cocoa; coffee; cotton; sugar None None Copra; coconut oil; cotton; sugar None None Coffee; cotton; sugar

90

Analysis of Movements in Productivity and Prices Niger Nigeria Papua New Guinea Paraguay Peru Rwanda Saudi Arabia Senegal Sierra Leone Solomon Islands South Africa Sri Lanka Sudan Swaziland Syria Tanzania Togo Tonga Trinidad and Tobago Tunisia Turkey Uganda Uruguay Vanuatu Venezuela Zambia Zimbabwe

Cotton Cocoa; cotton; palm oil; palm kernel oil Cocoa; coffee; copra; coconut oil; palm oil; palm kernel oil Coffee; cotton; palm kernel oil; sugar Coffee; cotton; sugar Coffee None Cotton Cocoa; coffee; palm kernel oil Cocoa; copra; coconut oil; palm oil Cotton; rice; sugar Copra; coconut oil; rice Cotton; sugar Cotton; sugar Cotton Cocoa; coffee; cotton, sugar Cocoa; coffee; cotton Copra; coconut oil Cocoa; sugar None Cotton; sugar Coffee; cotton Rice; sugar Cocoa; copra; coconut oil Cocoa; coffee; rice; sugar Coffee; cotton; sugar Coffee; cotton; sugar

reported for each country were not always produced consistently throughout the study period. Some that were important exports in the early years for a particular country but were not exported in the latter years of the period tend to be omitted.

4.6.2. Production 4.6.2.1. DESCRIPTION OF PRODUCTION DATA The decision on how many output categories to include in the analysis was dictated by the size of the data set and the number of commodity categories for which all countries would have non-zero output values. In theory, the number of outputs could have been as large as the number of commodities for which data were available, but we had to restrict the categories to aggregate crops and aggregate livestock to avoid zero outputs for a substantial number of countries. Output aggregates were obtained from Rao (1993, Table 5.4) and constructed using international 1990 prices denominated in US dollars. For most countries, both crop and livestock data are available for the study period. For some countries where livestock activities are relatively unimportant, however, separate livestock data were unavailable so TFP estimates were based on crop production only. The number of inputs was dictated by the availability of data. Five categories were included: land area, tractors, labour, fertilizer, and livestock (in dry sheep

91

The Issue of Declining Commodity Prices equivalents). While it would have been desirable to have tree inputs given our interest in a number of tree crops, data on the cost of seedlings were not available and, if they were, would not have been accurate given the common policy of subsidizing the dissemination of seedlings to farmers. In any event, the two most important inputs in the establishment and maintenance of tree crops are labour and fertilizer, which are both included in the input set. The proportions of the value of output of the selected tropical commodities to the total value of crop output in each country in the ﬁnal year of the study period, 2002, are presented in Table 4.4. They vary markedly from zero to almost all agricultural output. Those countries with zero or negligible percentages are seldom considered in the discussions that follow. Other countries with high percentages, such as Malaysia, are of less concern because most of the output of selected commodities (e.g. rice in the case of Malaysia) is for domestic consumption. Table 4.4. Proportion of the Total Value of Crop Output Contributed by the Selected Commodities in 1990 Country Algeria Angola Argentina Bangladesh Barbados Benin Bolivia Botswana Brazil Burkina Faso Burundi ˆ te d’Ivoire Co Cambodia Cameroon Central African Republic Chad Chile China Colombia Congo, Democratic Republic of Costa Rica Cuba Dominican Republic Ecuador Egypt El Salvador Fiji Islands Gabon Gambia, The Ghana Guatemala Guinea Guinea-Bissau Haiti

92

Proportion of the total value of crop output (%) 0.03 22.39 5.49 77.47 79.09 15.01 22.04 2.03 31.81 12.50 7.64 61.05 70.92 42.08 20.61 14.47 1.33 30.89 56.82 18.49 45.46 76.68 44.29 45.56 16.76 52.70 92.64 13.24 25.44 43.81 41.72 50.23 48.62 21.21

Analysis of Movements in Productivity and Prices Honduras India Indonesia Iran Jamaica Kenya Laos Lesotho Liberia Madagascar Malawi Malaysia Mali Mauritania Mauritius Mexico Mongolia Morocco Mozambique Namibia Nepal Nicaragua Niger Nigeria Papua New Guinea Paraguay Peru Rwanda Saudi Arabia Senegal Sierra Leone Solomon Islands South Africa Sri Lanka Sudan Swaziland Syria Tanzania Togo Tonga Trinidad and Tobago Tunisia Turkey Uganda Uruguay Vanuatu Venezuela Zambia Zimbabwe

45.85 37.77 67.09 6.73 30.73 17.01 67.25 0.00 44.31 50.98 7.77 90.25 22.64 33.40 84.86 16.96 0.00 0.99 11.33 0.06 46.46 42.72 3.34 23.91 48.74 19.07 34.20 5.56 0.00 13.30 68.88 78.40 10.35 52.10 17.22 61.59 0.00 15.07 25.98 19.28 70.54 0.01 2.30 7.08 21.52 83.20 29.61 9.44 13.18

4.6.2.2. MANIPULATION OF PRODUCTION DATA A Fisher index was constructed to obtain an agricultural output index for each country. The index was initially estimated for output data in 1990, following the approach adopted by Coelli and Rao (2004) and Alauddin et al. (2005). FAO production indices were then used to calculate crop and livestock output data for each year in each country back to 1970 and from 1991 onwards. Some

93

The Issue of Declining Commodity Prices but not all data were available for the years 2003 and 2004, which meant the study period had to end in 2002.

4.6.3. Commodity price data Price data are obtainable from the FAO website for more than 150 commodities at the producer, export and global levels. It would be a mammoth task to construct an aggregated commodity price series for every country under study so we conﬁned our attention to the eight commodities of special interest. Producer equivalent values were extracted for the exports of paddy rice equivalent (the broad rice category was used), coffee green beans, cocoa beans, raw sugar equivalent, cotton lint, copra, palm kernel oil and coconut oil (comprising the lauric oil category), and palm oil as the ﬁrst step to calculate export unit values. The export unit value index was then constructed by extracting export quantities for the same group of commodities, and dividing export value by export quantity for each commodity. Once the export unit value series was established for each commodity, an aggregate Fisher index was constructed for all commodities. A similar procedure was followed to construct the producer price index for these commodities, also using the FAO AGROSTAT data series. Again, a Fisher price index was constructed for each country. The FAO data series for producer prices is incomplete in every country, unlike the export unit value data. Therefore, most emphasis is placed on the latter series in undertaking the analyses in the next section. But the producer price data provide useful comparisons and some additional information on price trends that is used to construct single factoral terms of trade indices.

4.7. Results 4.7.1. Trends in commodity prices 4.7.1.1. EXPORT UNIT VALUES Indices of export unit values of the selected tropical commodities aggregated across all countries in the sample for the period 1970–2002 are summarized in Table 4.5. Annual rates of price decline with standard errors were estimated by taking the logarithm of the dependent variable, export unit values, and regressing it on a trend variable. Indices for individual countries are presented in Appendix 4.2. As with trends in export quantities, price trends are described separately for all selected commodities, tree crops (coffee, cocoa, palm oil and lauric oils) and ﬁeld crops (rice, cotton and sugar). Separate price trend analyses were undertaken for Commonwealth countries, African countries and African Commonwealth countries for the same period from 1970 to 2002.

94

Analysis of Movements in Productivity and Prices Table 4.5. Estimates of Trends in Export Unit Values of Selected Commodities, 1970 to 2002 Variable All countries Total commodities: Intercept Trend R square Tree crops: Intercept Trend R square Field crops: Intercept Trend R square Commonwealth countries Total commodities: Intercept Trend R square Tree crops: Intercept Trend R square Field crops: Intercept Trend R square African countries Total commodities: Intercept Trend R square Tree crops: Intercept Trend R square Field crops: Intercept Trend R square

Coefﬁcient

Standard error

t-statistic

P-value

1.172 0.032 0.731

0.096 0.003

12.265 9.176 Standard error

1.98E13 2.4E10 0.189

1.454 0.032 0.543

0.144 0.005

10.10 6.071 Standard error

2.51E11 1E06 0.284

0.906 0.035 0.847

0.073 0.003

12.405 13.108 Standard error

1.48E13 3.48E14 0.144

1.182 0.028 0.678

0.095 0.003

12.499 8.071 Standard error

1.21E13 4.1E09 0.187

1.425 0.030 0.574

0.129 0.005

11.073 6.467 Standard error

2.66E12 3.28E07 0.254

0.655 0.022 0.677

0.075 0.003

8.785 8.058 Standard error

6.43E10 4.23E09 0.147

1.282 0.033 0.715

0.104 0.004

12.295 8.825 Standard error

1.86E13 5.8E10 0.206

1.571 0.037 0.595

0.153 0.006

10.252 6.755 Standard error

1.76E11 1.46E07 0.303

0.673 0.023 0.681

0.078 0.003

8.655 8.131 Standard error

8.96E10 3.49E09 0.154

0.102 0.004

11.664 8.199 Standard error

7.19E13 2.92E09 0.299

0.151 0.005

9.931 6.454 Standard error

3.78E11 3.39E07 0.299

0.104 0.004

5.451 4.918 Standard error

5.9E06 2.72E05 0.206

African Commonwealth countries Total commodities: Intercept 1.193 Trend 0.030 R square 0.573 Tree crops: Intercept 1.502 Trend 0.035 R square 0.573 Field crops: Intercept 0.568 Trend 0.019 R square 0.438

95

The Issue of Declining Commodity Prices Export unit values of the selected tropical commodities for all countries Export unit values of the selected tropical commodities for all countries in the sample for the period from 1970 to 2002 are presented in Figure 4.6. Linear trend lines are also included for each series. Clear evidence is presented of signiﬁcantly declining trends in both groups of commodities over the study period. The average annual price decline for all commodities is 3.17 per cent (Table 4.5). The average annual price decline for tree crops of 3.15 per cent is slightly lower than that for ﬁeld crops at 3.46 per cent. Export unit values of the selected tropical commodities for Commonwealth countries Export unit values of the selected tropical commodities for Commonwealth countries in the sample are presented in Figure 4.7. The annual rate of price decline for all commodities in Commonwealth countries is 2.76 per cent (Table 4.5), which is slightly lower than for all countries in the sample because of a lower rate of decline for ﬁeld crops (2.17 per cent per annum). The annual rate of decline for tree crops (3.01 per cent) is close to the rate for the total sample, reported above. Export unit values of the selected tropical commodities for African countries Export unit values of all selected tropical commodities, selected tree crops and selected ﬁeld crops are presented in Figure 4.8 for all African countries in the sample. The overall rate of price decline of 3.32 per cent corresponds closely to that for all countries but, as for the Commonwealth countries, the annual rate of price decline for ﬁeld crops of 2.28 per cent is quite a bit lower than the rate for all countries of 3.46 per cent. Export Price Index - All Tropical Countries (Deflated) 6 Total price index Tree crop price index

5

Index (1990=1.0)

Field crop price index Linear (Tree crop price index)

4

Linear (Total price index) Linear (Field crop price index)

3

2

1

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

0

Year

Figure 4.6. Export Price Index for Selected Commodities in All Sampled Countries, 1970–2002

96

Analysis of Movements in Productivity and Prices Export Price Index - Commonwealth Countries (Deflated) 4.5 Total price index

4

Tree crop price index Field crop price index

Index (1990=1.0)

3.5

Linear (Tree crop price index)

3

Linear (Total price index) Linear (Field crop price index)

2.5 2 1.5 1 0.5 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

0

Year Figure 4.7. Export Price Index for Selected Commodities in Sampled Commonwealth Countries, 1970–2002

Export Price Index - Africa (Deflated) 6 Total price index Tree crop price index Field crop price index Linear (Total price index) Linear (Field crop price index) Linear (Tree crop price index)

Index (1990=1.0)

5

4

3

2

1

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

1984

1983

1982

1981

1980

1979

1978

1977

1976

1975

1974

1973

1972

1971

1970

0

Year Figure 4.8. Export Price Index for Selected Commodities in Sampled African Countries, 1970–2002

Export unit values of the selected tropical commodities for African Commonwealth countries Finally, export unit values of the selected tropical commodities are presented in Figure 4.9 for African Commonwealth countries in the sample. Again, the rate of price decline is substantially lower for ﬁeld crops (1.85 per cent per annum)

97

The Issue of Declining Commodity Prices Export Price Index - Commonwealth Africa (Deflated) 6 Total price index Tree crop price index Field crop price index Linear (Tree crop price index) Linear (Total price index) Linear (Field crop price index)

Index (1990=1.0)

5

4

3

2

1

1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

0

Year Figure 4.9. Export Price Index for Selected Commodities in Sampled African Commonwealth Countries, 1961–2002

than for tree crops (3.53 per cent). The higher rate of decline for the tree crop index is dictated largely by successively smaller magnitudes of price rises during commodity booms following the major boom of the mid-1970s. Relationship between world import unit values and export unit values of the selected tropical commodities for all countries The share that farmers receive of the landed value of their exports in the importing countries can change in two domains: from the farm gate to the point of export and between the point of export and point of import. The exporters’ share in the latter domain can change as a result of changes in the technology and costs of commodity transfer, and changes in bargaining power between exporters and importers. In this section, we examine trends in this second domain to ascertain whether free-on-board export prices of commodities decreased more quickly or more slowly than landed import prices over the study period. If they decreased more quickly, it indicates that the sampled countries have suffered a relative decline in their share of the landed value of the selected commodities at the point of import. Indices of the export unit values and the world import unit values of the selected tropical commodities for the study period are presented in Figure 4.10 for all sampled countries. No divergence in trend between the two indices is apparent: if anything import unit values have declined more than the export unit values, but the divergence is not statistically signiﬁcant. The relative trends in export unit values and world import unit values can be examined separately for tree crops and ﬁeld crops. First, Figure 4.11 illustrates these trends for tree crops. As observed above for crops in aggregate, the ratios

98

Analysis of Movements in Productivity and Prices Total Commodity World Import Price and Export Price Index - All Sampled Countries 5.0 Export price index World import price index Linear (World import price index) Linear (Export price index)

Index (1990=1.0)

4.0

3.0

2.0

1.0

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

1984

1983

1982

1981

1980

1979

1978

1977

1976

1975

1974

1973

1972

1971

1970

0.0

Figure 4.10. Export Price and Import Price Indices for All Commodities in All Sampled Countries, 1970–2002 5.0

Index (1990=1.0)

4.0

Tree Crops World Import Price and Export Price Indices - All Sampled Countries World import price index Tree crops export price index Linear (Tree crops export price index) Linear (World import price index)

3.0

2.0

1.0

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

0.0

Figure 4.11. Export Price and Import Price Indices for Tree Crops in All Sampled Countries, 1970–2002

of export unit values to the world prices of the selected tree crop commodities for the study period track each other very closely. Again, there is no statistically signiﬁcant divergence between the two indices. Second, Figure 4.12 illustrates the trends in indices for ﬁeld crops. The respective two indices for the selected ﬁeld crop commodities do not track as closely as do those for the tree crop commodities. But the trend lines are virtually parallel

99

The Issue of Declining Commodity Prices Field Crops World Import Price and Export Price Indices - All Sampled Countries 2.5 World import price index Field crops export price index

Index (1990=1.0)

2.0

Linear (Field crops export price index) Linear (World import price index)

1.5

1.0

0.0

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

0.5

Figure 4.12. Export Price and Import Price Indices for Field Crops in All Sampled Countries, 1970–2002

and yet again there is no statistically signiﬁcant divergence between them. It is concluded that there is no evidence to support the proposition that the sampled countries have suffered a relative decline in their share of the landed value of the selected commodities between the point of export and the point of import. 4.7.1.2. PRODUCER PRICES An index of producer prices of the selected tropical commodities was estimated for each country. Because the time period for the series varies from one country to the next, there is little to be gained in estimating aggregate price series, and their trends, as done above for export unit values. A data ﬁle and details of the producer price trend for each country are available on demand. Trends in producer prices for individual countries are not always consistent with those for export unit values. Exogenous factors such as extreme climatic variations can lead domestic prices to vary from world prices but more often than not there are systemic factors at work in the domestic agricultural economies. Hertel and Winters (2005), Winters, McCulloch, and McKay (2004), Nicita (2005) and Von Braun, Boue¨t, Cororaton, Mengistu, and Orden (2005) identiﬁed the main sources of discrepancy between border prices and producer prices arising from imperfect price transmission as: . variations in infrastructure and transport costs . market imperfections . domestic ﬁscal policies and regulations.

100

Analysis of Movements in Productivity and Prices Trends in the two series for each country were checked to see if they display reasonable correspondence. Evidence for 54 countries, covering varied time periods between 1970 and 2002, provides little evidence of similar trend magnitudes in indices of export unit values and producer prices. In some cases, even the direction of change differs. This overall result suggests that there are major obstacles to the effective transmission of commodity prices back along the value chain from the point of export to the point of agricultural production. Producer prices show a declining trend over the different estimation periods in 40 countries, compared with 48 countries for which a declining trend is evident in export unit values. In 34 countries, the rate of decline was lower, or the rate of increase higher, in producer prices than in the corresponding rate of change in export unit values. In 26 of these countries, the difference in magnitude was greater than one per cent. For example, the producer price index increased strongly relative to export unit value index in Ghana. This result accords with the ﬁnding by Choraria (2005), based on commodityspeciﬁc and rigorous analysis, that Ghanaian cocoa producers were capturing a higher proportion of the export dollar over time. The magnitude of the lower rate of decline, or higher rate of increase, in export unit values over producer prices was greater than one per cent in twelve countries of the twenty countries in which the ratio of export unit values to producer prices increased. Some other estimates in this study ﬁt closely the results Choraria (2005) reported in her analysis of compression in speciﬁc commodity export supply chains in a number of developing countries. Choraria (2005) detected no signiﬁcant trend in the coffee value chain between the point of export and the farm gate in Cameroon or Kenya, despite a signiﬁcant decline in the producer to retail price ratio. The same results were arrived at in this study for the indices of all selected export commodities in those countries. Similarly, no signiﬁcant trend was found in Mauritius for either the ratio of producer to export price of sugar or for the index of selected commodities of that country.

4.7.2. Trends in productivity in agriculture 4.7.2.1. LABOUR PRODUCTIVITY CHANGE Three-year moving averages were estimated of the change in labour productivity in the agricultural sector by country for the period from 1970 to 2001. A data ﬁle and details of estimates of the labour productivity trend for each country are available on demand. Figure 4.13 shows the labour productivity index for 83 developing countries from the base year of 1970 (labour productivity index ¼ 1.0) until the ﬁnal year of the study period, 2002. Twenty-seven per cent of the countries (those to the left of the red vertical line) did not achieve any growth in labour productivity during the period and 43 per cent

101

The Issue of Declining Commodity Prices 16 14 12 10 8 6 4 2 0 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 More

2002 Labour productivity index (1970 = 1.0) Figure 4.13. Annual Rates of Change in Labour Productivity in All Sampled Countries, 1970–2002

experienced and achieved a cumulative labour productivity growth rate of less than 20 per cent, which translates to an annual labour productivity growth rate of less than 0.6 per cent. Those countries to the left of the dotted line (88 per cent of the sample) increased their labour productivity by no more than 3 per cent per annum. 4.7.2.2. TOTAL FACTOR PRODUCTIVITY CHANGE Three-year moving averages were estimated of the change in TFP in the agricultural sector by country for the period from 1961 to 2001. In one country, Burundi, TFP estimates proved unreliable due to the erratic recording of data on tractors and fertilizers. Labour productivity estimates were used as a proxy for TFP in this case. As these inputs tended to increase at around the same rate as labour over the study period, labour productivity was considered to be a reasonable approximation to TFP. Figure 4.14 shows the cumulative TFP index for 83 developing countries from the base year of 1970 (TFP index ¼ 1.0) until 2002. Forty-one per cent of the countries (those to the left of the solid vertical line) did not experience overall TFP growth during the period and slightly more than half achieved a cumulative TFP growth rate of less than 20 per cent, which translates to an annual TFP growth rate of less than 0.6 per cent. The TFP of those countries to the left of the dotted line (90 per cent of the sample) less than doubled TFP over the study period, which means their annual TFP growth rate was less than 2.2 per cent per annum.

102

Analysis of Movements in Productivity and Prices Many countries (29, or 35 per cent of the sample) experienced declines in TFP between 1986 and 2002, a period when the fruits of scientiﬁc endeavour and improved economic policies would have been expected to be enjoyed through progress in agricultural production technology and productivity growth. A handful of these countries might have suffered productivity setbacks as a result of civil wars and extended periods of drought. A disturbingly high number of countries in this category are classiﬁed as least developed countries. At least as many others enjoyed productivity gains simply by catching up with technology adoption that had been delayed by the same factors earlier in the study period. Indonesia is a case in point showing productivity decline despite TFP gains over the whole period. By 1985, its cumulative TFP index had reached 1.72, an impressive 72 per cent increase over the TFP level in 1970. Yet the cumulative TFP index had declined to 1.65 almost two decades later in 2002. Fuglie (2004) discussed the events surrounding this trend and concluded that stagnation in TFP growth threatens to undermine government efforts to bring about widespread rural development. Chile achieved the highest TFP growth rate at 4.27 per cent per annum. A feature of its growth was the sharp distinction in agricultural performance between the 1970s and 1980s on one hand and the 1990s on the other. In 1989, the cumulative TFP index stood at 1.10, a total increase in TFP of just 10 per cent over the ﬁrst two decades. By 2002, the cumulative index had risen extraordinarily rapidly to 3.81, a result no doubt of the fruits of technological progress but also an indication of the success achieved in moving agricultural production into more intensive higher value-adding industries such as deciduous fruits. Around 70 per cent of the TFP change in Chile was due to technological change and around 30 per cent to efﬁciency improvement. 18 16 14 12 10 8 6 4 2 0 0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

2.6

2.8

3.0 More

2002 TFP (1970 = 1.0) Figure 4.14. Annual Rates of Change in TFP in All Sampled Countries, 1970–2002

103

The Issue of Declining Commodity Prices 4.7.2.3. COMPARISON OF LABOUR PRODUCTIVITY CHANGE AND TFP CHANGE Average annual labour productivity and TFP estimates by country over the study period are presented in Appendix 4.3. There is a wide variation in trends across the sampled countries for both measures. But a comparison between these two sets of estimates reveals that twice as many countries achieved higher growth (or lower decline) in labour productivity than in TFP as achieved higher growth (or lower decline) in TFP than in labour productivity. This result is not surprising given the expected movement of people out of agriculture into the secondary and tertiary sectors over the 32-year period and the substitution of capital (especially machinery) for labour in agricultural production. Higher rates of labour productivity growth are particularly noticeable in countries achieving substantial economic growth, such as Malaysia. Annual growth in labour productivity in Malaysia was high, at 4.27 per cent, while annual TFP growth was much lower, at only 0.33 per cent. The number of economically active people in agriculture declined by almost 9 per cent between 1970 and 2002 while all other inputs increased substantially, especially machinery use. Less spectacular but still substantial differences between labour productivity and TFP growth rates were experienced in many African countries. Interestingly, former French colonies tended to have substantially higher annual growth rates in labour productivity than TFP. This list includes Benin (3.64 per cent versus 1.49 per cent), Burkina Faso (1.86 per cent versus 1.90 per ˆ te d’Ivoire (1.68 per cent versus 0.68 per cent), Chad (0.73 per cent cent), Co versus 0.69 per cent), Gabon (3.34 per cent versus 2.01 per cent), Mali (1.79 per cent versus 1.37 per cent), Mauritania (1.30 per cent versus 0.27 per cent), Niger (0.25 versus 1.28) and Togo (0.53 versus 0.17). On the other hand, most former British colonies tended to have substantially higher annual growth rates in labour productivity than TFP. One country in the latter group, Kenya, and two Asian countries, Bangladesh and Nepal, provide an interesting contrast to the example of Malaysia given above. Economic growth took place in these countries but a combination of growth in intensiﬁcation of agricultural production, high population growth in rural areas and limited off-farm employment opportunities meant that the number of economically active people in agriculture increased substantially over the period (by 163 per cent in Kenya, 84 per cent in Nepal and 44 per cent in Bangladesh). Kenya’s high rate of increase in rural labour resources resulted from an extremely high population growth rate approaching 4 per cent, such that ‘by the late 1970s, Kenya was beginning to acquire some of the characteristics of a labour-surplus economy’ (Perkins et al. 2001, p. 98). Bangladesh achieved only a small increase in

104

Analysis of Movements in Productivity and Prices labour productivity, at 0.86 per cent per annum, compared with an impressive annual TFP growth rate of 3.34 per cent. Roberts and Fagerna¨s (2004, p. 12) reported that ‘Bangladesh’s long stagnant farm sector experienced accelerating growth starting in the later 1980s, due in large part to the widespread and successful implementation of Green Revolution technology, accompanied by the liberalization of prices, input supply and marketing’. Equivalent rates for Nepal were 1.16 and 1.99. Events in Nepal closely followed those in Bangladesh with its farmers also coming later to technological change of the Green Revolution type. Nepal’s cumulative TFP index was still below unity in 1979. In Kenya, TFP growth was a respectable but moderate 1.16 per cent per annum, in large part due to the expansion of horticultural exports that have accounted for two-thirds of all growth in agricultural exports (World Bank 2003). Annual labour productivity growth was substantially lower at only 0.14 per cent, in part a reﬂection of the high labour intensity in horticultural production. Labour absorption in agriculture in Bangladesh has slowed down in recent years. An analysis by Alauddin and Tisdell (1995) of labour productivity and TFP growth rates in Bangladesh until the early 1990s led them to comment that ‘there is little prospect of agriculture providing much greater employment’ (Alauddin and Tisdell 1995, p. 281). Indeed, TFP grew strongly by 56 per cent over the seven years after their analysis was published, yet employment in agriculture increased by only six per cent over the same period. Alauddin and Tisdell (1995, p. 281) concluded that ‘there is little sign that labor transfer in Bangladesh will lead to a pattern of successful industrialization’. Chile provides yet another example of a country that has managed to achieve an annual TFP growth (4.27 per cent) well in excess of annual growth in labour productivity (2.01 per cent). Between 1970 and 2002, its large neighbours, Brazil and Argentina, experienced declines in the number of economically active people in agriculture of 21 per cent and two per cent, respectively, and had annual labour productivity growth rates well in excess of annual TFP growth rates. In contrast, Chilean agriculture experienced a growth in its labour force of 38 per cent over the same period. 4.7.2.4. COMPARISON OF RATES OF CHANGE IN TFP BETWEEN COMMONWEALTH COUNTRIES Figure 4.15 contains a summary of the rates of change in TFP across 29 Commonwealth countries in the sample. Thirty-eight per cent of these countries failed to make progress in TFP change over the study period and 31 per cent suffered rates of TFP decline greater than 1 per cent per annum. As mentioned above, Bangladesh achieved a rapid TFP growth rate, which was highest among Commonwealth countries and second highest behind Chile for the whole sample. This high growth rate derived from technological

105

The Issue of Declining Commodity Prices 12 10 8 6 4 2 0

−2.0

−1.0

0.0

1.0

2.0

3.0

4.0

Annual change in TFP (%) Figure 4.15. Annual Rates of Change in TFP in Commonwealth Countries, 1970–2002

change rather than change in technical efﬁciency. It reﬂects the later start by its farmers in reaping Green Revolution rewards compared with farmers in some other Asian countries that beneﬁted from these technology gains, such as India. The cumulative TFP index was below unity throughout the 1970s and surpassed this ﬁgure for the ﬁrst time in 1980. India achieved a moderate annual TFP growth of 1.44 per cent, which was evenly spread over the study period. By 1983, India’s cumulative TFP index had reached 1.20 compared with 1.08 for Bangladesh. Only three countries (10 per cent of the Commonwealth sub-sample) attained TFP growth rates above 2 per cent per annum. All were located in Southern Africa: South Africa and two agricultural economies closely linked to its sector, Swaziland and Namibia, where technical efﬁciency change was more prominent than technological change. TFP progress in South Africa was almost entirely due to technological change. TFP growth rates were between one per cent and two per cent per annum in four countries, two of which are mentioned above (India and Kenya), where TFP growth was generated predominantly by technological progress. Nigeria experienced a growth rate very close to 2 per cent, with technological change again the dominant force. Its performance can be divided into two periods. From 1970, TFP slowly sunk until the mid-1980s in the wake of the ‘Dutch disease’ inﬂuence of petroleum exports and an inclement economic environment. The agricultural sector, however, has been the beneﬁciary of an extensive research and extension system that has produced some signiﬁcant on-farm productivity gains. Despite some glaring shortcomings in the system, as outlined by Okunmadewa and Olayemi (2002), there have been a number of active programs, augmented by the location of some international research scientists in the country.

106

Analysis of Movements in Productivity and Prices A small South Paciﬁc country, Solomon Islands, just reached the threshold of 1 per cent TFP growth rate. Its index was boosted in large part by technological change associated with the successful development of an oil palm industry from the early 1970s until 1986 when a cyclone decimated the agricultural sector, including the oil palm industry. In that year, the cumulative TFP index had reached 1.40, a level that remained above the 2002 index of 1.38. As pointed out above, several countries experienced declines in TFP from the mid-1980s. A number of Commonwealth countries that experienced annual productivity growth rates for the whole study period of less than one per cent fall into this category. They include Jamaica, Mauritius, Sri Lanka, Tanzania and Uganda. IADB (2003, pp. 19–26) pinpointed many reasons for low levels of growth in TFP and labour productivity in the general economy in Jamaica. Three factors were mentioned that have particular relevance to the agricultural sector: lack of use of modern technology, low standards of education and training, and weak infrastructure. TFP trends in Uganda largely mirror the fortunes of the coffee industry, with TFP gains in the 1960s and 1970s turning to a slight decline thereafter. APEP (2005, p. 1) reported that several factors had contributed to a decline in productivity in the coffee industry, including ‘Diseases and pests, notably coffee wilt disease . . . Old coffee trees . . . Poor crop management practices . . . Poor soil fertility management [and] Poor post-harvest handling practices’. Malaysia and Papua New Guinea narrowly escape appearing on this list but their TFP growth has been very low over the past two decades. In the case of Malaysia, the TFP index in 1982 was fractionally higher than the index in 2002 (the 1985 ﬁgure was actually slightly lower). In Papua New Guinea, very slow TFP growth over most of the study period turned into decline in 2001. Two countries mentioned above in groups with higher TFP growth, Solomon Islands and Swaziland, also fall into the category of TFP decline since the mid-1980s. The eleven Commonwealth countries that experienced a decline in their TFP index during the whole study period (with percentage annual decline in parentheses) were: Ghana (0.24); Cameroon (0.92); Tonga (1.02); Sierra Leone (1.04); Zambia (1.21); Vanuatu (1.27); Mozambique (1.47); The Gambia (1.68); Barbados (2.18); Fiji (2.47); and Trinidad and Tobago (2.53). The conspicuous presence of ﬁve small island countries among this group suggests they are facing particular difﬁculties in generating TFP growth (or even maintaining existing levels) among traditional export industries on their limited land areas. Two other countries in this group—Sierra Leone and Mozambique—experienced periods of civil war during the study period. In the case of Sierra Leone, this period was towards the end of the study period and its effects are shown by the decline in the cumulative TFP index from 0.94 in 1998 to 0.72 in 2002.

107

The Issue of Declining Commodity Prices Mozambique experienced protracted conditions of civil war earlier in the study period. The cumulative TFP index plummeted from 1.0 in 1970 to 0.47 in 1994 when it reached its nadir. Recovery has been steady since that date following the implementation of a structural adjustment program following the cessation of hostilities in 1992. Recovery was delayed by a bad drought that coincided with peace (Arndt 2005, p. 3). The index averaged 0.66 in the ﬁnal ﬁve years of the study period, but declined slightly in the ﬁnal two years. Ghana and Cameroon present interesting cases that illustrate the need to scrutinize trends for break points over the 32-year study period. From the start of the period (and indeed back to the early 1960s), Ghana experienced substantial declines in TFP in its major cash crop industries due to gross economic mismanagement. The cumulative index slumped to 0.48 in 1983, less than half the index in 1970. The situation began to improve from the mid-1980s, a trend clearly seen from the fact that the cumulative TFP index more than doubled to 0.99 by 2000. Ominously, TFP decline re-emerged in 2001 and the index closed at 0.93 in 2002. TFP performance in Cameroon between 1970 and 2002 closely follows the macroeconomic policy adopted by the government. Amin, Douya and Mbeaoh (2002, p. 155) characterized government policy until the mid-1980s as ‘a protectionist policy [that] was combined with state intervention in all spheres of the economy and strict price controls that effectively prohibited the development of a viable market system’. Despite some impressive growth statistics during this period, TFP in the agricultural sector almost halved between 1970 and 1985. The adverse effects on TFP of this economic policy were augmented, ﬁrst, in the early 1980s by the presence of ‘Dutch disease’ effects induced by the petroleum industry that caused stagnation in the agricultural sector and, second, by a ﬁnancial crisis that caused economic decline between 1985 and 1988 (Amin, Douya, and Mbeaoh 2002, p. 155). The implementation of economic reforms from 1988 entailing ‘liberalization of trade in agricultural products; privatization of agricultural production and dissolution of state-owned agroindustrial corporations’ (Amin, Douya and Mbeaoh 2002, p. 157) coincided with a noticeable revival in TFP, which grew continuously from 1989 to 2002. The cumulative TFP index in 2002 was 49 per cent higher than the 1989 level but still below the level in 1970. Other Commonwealth countries experiencing TFP decline over the whole study period also experienced decline from 1985 to 2002. Included in this group are Barbados, Fiji, The Gambia, Sierra Leone, Tonga, Vanuatu, and Zambia. The precipitous decline in the cumulative TFP index in Fiji, from 0.86 in 1991 to 0.45 in 2002, is attributable to a combination of civil strife that spilt over into rural areas affecting the Indian population that provided most of the highly productive farmers, land tenure problems affecting Indian tenant farmers (Kurer 2001), and declining soil fertility. All of these factors have had especially adverse impacts on production in the dominant sugar industry.

108

Analysis of Movements in Productivity and Prices With their heavy reliance on cotton as a source of cash income from exports, many smallholders in Zambia suffered from a moribund industry until reforms were instituted in 1994 (Balat and Porto 2005). By this time, the cumulative TFP index had slumped to 0.65 from 1.17 in 1978. Reforms were implemented in the cotton sector in 1994 as part of a market liberalization program but, as Balat and Porto (2005) observed, they experienced difﬁculties. Although TFP showed signs of improvement in the latter part of the 1990s, the cumulative TFP index stood at only 0.68 in 2002. Among non-Commonwealth African countries, Burkina Faso exempliﬁes the problems many countries have had in maintaining, let alone improving, TFP. Its agricultural sector suffered an annual decline in TFP of 1.9 per cent over the study period, with desertiﬁcation and locust plagues hindering efforts to raise TFP. Henao and Baanante (1999, p. 2) reported that it ‘would have to increase its NPK consumption more than 11 times to maintain crop production levels without depleting nutrients’.

4.7.3. Explaining changes in productivity in developing agriculture 4.7.3.1. FACTORS EXPLAINING CHANGES IN TFP The estimated model of factors explaining changes in TFP is presented in Table 4.6. Continuous explanatory variables and the dependent variable, TFP, were converted into natural logarithms so that their coefﬁcients could be directly interpreted as elasticities. The ﬁrst point to note is that deﬂated commodity prices have a signiﬁcant negative effect on TFP, implying that producers increase their TFP growth rate when confronted by an increase in the rate at which real commodity prices fall. But the extent of this response is extremely limited, as indicated by the elasticity: a one per cent decline in the commodity price index leads to only a 0.007 per cent increase in the TFP growth rate. This result stands to reason in that producers usually face major technical and knowledge barriers to increasing their TFP in commodity production in the short term. Separate regression analyses were undertaken for 13 Commonwealth and African countries, including only explanatory variables that change from year to year but without the variables of proportions of selected commodities. They show a divergence in estimates, indicating that producers in many countries do not respond signiﬁcantly to changes in the commodity price index. All countries that do record signiﬁcant negative responses have low elasticities of the order reported above. The second result of particular interest is the different rates at which TFP changes according to the relative importance of the selected commodities in the product mix of the agricultural sector in each country. Here, results accord closely to prior expectations. An increase in tree crops in the product mix

109

The Issue of Declining Commodity Prices Table 4.6. Estimated TFP Model Variable

Name

Deﬂated price index GDPPC85 Labour productivity 1970 Rural proportion Illiteracy rate Proportion of cocoa Proportion of coconut Proportion of coffee Proportion of palm oil Proportion of rice Proportion of cotton Proportion of sugar Latin America Africa Rain Intercept

PIDC GDPPC85 CONV RURAL ILLIT PROPF1 PROPF2 PROPF3 PROPF4 PROPF5 PROPF6 PROPF7 DRL DRA RAIN CONSTANT

Estimated coefﬁcients

Standard error

0.0066 0.0006 0.0071 0.0024 0.0076 0.0817 0.1235 0.0287 0.0443 0.0131 0.1420 0.0063 0.0319 0.0113 0.0058 0.0703

0.0029 0.0050 0.0055 0.0083 0.0043 0.0482 0.0549 0.0283 0.0195 0.0156 0.0906 0.0249 0.0094 0.0058 0.0034 0.0895

t-ratio

p-value

2.300 0.115 1.290 0.284 1.749 1.693 2.249 1.014 2.279 0.839 1.567 0.252 3.404 1.969 1.707 0.785

0.022 0.908 0.197 0.777 0.08 0.091 0.025 0.311 0.023 0.401 0.117 0.801 0.001 0.049 0.088 0.432

results in lower rates of TFP growth. The only exception to this result is coffee production where the TFP growth rate does not signiﬁcantly alter for a change in the proportion of coffee output in the product mix. Elasticities are very low for all forms of commodity production. Of the ﬁeld crops, sugar production was expected to show a result similar to that for tree crops. That is, a higher proportion of sugar in the agricultural output mix was thought to be associated with a lower rate of productivity growth. There are two reasons for expecting this result. The ﬁrst is an absence of technology spillovers from international research centres and the developed world compared with the other selected ﬁeld crops. Second, the land on which sugar is grown has been showing distinct signs of degradation in recent times in many sugar-producing countries. The estimated coefﬁcient does not support this proposition. Like coffee production, the proportion of sugar in the product mix has no signiﬁcant effect on the rate of TFP growth. A possible explanation of this result is that sugar is such a dominant crop in many of the countries where it is exported that the overall agricultural productivity index largely reﬂects the sugar productivity index. Both cotton production and rice production in less developed countries should have beneﬁted from substantial international research gains and their proportions in the agricultural product mix were expected to show signiﬁcant and positive coefﬁcients. Indeed, both coefﬁcients are positive but neither was signiﬁcant at the ten per cent signiﬁcance level (although the estimated coefﬁcient for cotton only narrowly fails this test). In conclusion, it appears safe to conclude that only the tree crops other than coffee have TFP growth rates substantially different from the average rates estimated for each country. While this result might not be valid for some countries, the low elasticities suggest that assuming average TFP growth rates

110

Analysis of Movements in Productivity and Prices for coffee, sugar, rice and cotton production is a sound way to proceed. But lower than average growth rates for cocoa, coconut and oil palm production mean that caution should be taken in inferring higher TFP growth rates than price declines for any country in which these commodities comprise a major proportion of agricultural output. On the other hand, they strengthen any ﬁnding for these countries of lower TFP growth rates than price declines. There are some interesting results for other explanatory variables. In line with expectations, the coefﬁcient on the illiteracy rate variable is negative and signiﬁcant with a p-value of 0.08. Long-term rainfall also has the expected positive sign with a p-value of 0.09. Rates of TFP change are signiﬁcantly lower in Latin America and Africa than in Asia. The proportion of rural population was found to have no signiﬁcant effect on TFP change. Similarly, against expectations, changes in GDP per head appear to have no signiﬁcant effect on TFP across the whole sample. However, this result varies widely between countries for individual analyses of the thirteen countries of interest. Most countries repeat the insigniﬁcant result but poorer countries with high proportions of rural population such as Bangladesh and Malawi had signiﬁcant and positive coefﬁcients. Countries in which growth in GDP per head was driven largely by growth in mineral and oil production where the beneﬁts of growth were not widespread, such as Papua New Guinea and Nigeria, have signiﬁcant and negative coefﬁcients, while less poor countries such as Ghana also recorded a signiﬁcant and negative coefﬁcient. The openness index, measured by the sum of export and import values as a proportion of GDP, could not be included in the general analysis, the results of which are reported in Table 4.6, because data were unavailable in many countries. It could be included in the individual analyses for the thirteen countries of interest, however, and provides mixed results. Some countries, especially those in which selected commodities form a substantial proportion of total exports, have a coefﬁcient on the openness index with at ˆ te d’Ivoire, least a reasonably signiﬁcant and positive sign (for example, Co Ghana and Jamaica). Others have signiﬁcant and negative coefﬁcients (for example, Malawi and Zambia). Yet others have insigniﬁcant coefﬁcients (for example, Kenya, Nigeria and Papua New Guinea). 4.7.3.2. FACTORS EXPLAINING CHANGES IN LABOUR PRODUCTIVITY The estimated model of factors explaining changes in labour productivity is presented in Table 4.7. Because our concern is with TFP rather than labour productivity, these results are included mainly for comparative purposes to conﬁrm the TFP results. Coefﬁcients on the variables representing inputs per labour unit as part of the Solow growth model are signiﬁcant and of expected positive sign.

111

The Issue of Declining Commodity Prices Table 4.7. Estimated Labour Productivity Model Variables

Name

Fertiliser/labour unit Machinery/labour unit Land/labour unit Deﬂated price index GDPPC85 Labour productivity 1970 Soil quality Rural proportion Illiteracy rate Proportion of cocoa Proportion of coconut Proportion of coffee Proportion of palm oil Proportion of rice Proportion of cotton Proportion of sugar Latin America South Asia Africa Rain Intercept

FERTPU TRACTPU LANDPU PIDC GDPPC85 CONV SOIL RURAL ILLIT PROPF1 PROPF2 PROPF3 PROPF4 PROPF5 PROPF6 PROPF7 DR1 DR2 DR4 RAIN CONSTANT

Estimated coefﬁcient

Standard error

0.0038 0.0107 0.1115 0.0099 0.1464 0.7788 0.0091 0.0323 0.1645 0.0326 0.5555 0.0686 0.1016 0.0192 0.4421 0.1648 0.2060 0.1240 0.1761 0.0898 2.2615

0.0014 0.0058 0.0176 0.0032 0.0108 0.0346 0.0021 0.0187 0.0177 0.1054 0.1378 0.0567 0.0527 0.0545 0.1578 0.0825 0.0464 0.0438 0.0394 0.0265 0.4255

t-ratio

p-value

2.739 1.836 6.345 3.133 13.510 22.510 4.245 1.722 9.292 0.310 4.031 1.21 1.929 0.352 2.801 1.997 4.439 2.831 4.472 3.388 5.315

0.006 0.066 0.000 0.002 0.000 0.000 0.000 0.085 0.000 0.757 0.000 0.226 0.054 0.725 0.005 0.046 0.000 0.005 0.000 0.001 0.000

The estimates generally conform to those discussed above for the TFP model but a few differences can be observed. First, the proportions of sugar and cotton in the product mix now accord more closely to expectations, with the proportion of sugar having a negative coefﬁcient and the proportion of cotton having a positive coefﬁcient, both with reasonably high signiﬁcance levels. The estimated coefﬁcient for the other variable of particular concern in this study, the deﬂated price index, has a slightly higher negative value but the estimate is little different from that in the TFP model. The coefﬁcient of the GDP per head variable is now highly signiﬁcant and positive. This result is to be expected: countries that have achieved a higher level of economic development have drawn more people from agriculture to meet increased labour demand in non-agricultural industries, resulting in greater substitution of labour-saving inputs in agricultural production. Hence, labour productivity is more closely associated with overall economic growth than is TFP. The coefﬁcient on the convergence variable is again positive but this time of higher magnitude and highly signiﬁcant. This result reinforces the ﬁnding of productivity divergence reported above. It reﬂects to some extent an inability of agricultural export industries in least developed countries to keep up with productivity growth rates achieved in less developed countries achieving higher rates of economic growth. Finally, a signiﬁcant but negative coefﬁcient is observed on the soil quality variable whereas a positive coefﬁcient was expected. (The coefﬁcient on this

112

Analysis of Movements in Productivity and Prices variable was insigniﬁcant in the estimation of the TFP model and it is excluded from the ﬁnal model.) It was thought that productivity growth would be higher in countries better endowed with soil resources. Failure to obtain such a relationship could be due to higher population densities in agricultural regions with better quality soils, resulting in more intensive labour use on farms in these areas. But this explanation is speculative and the causes of this relationship need further research.

4.8. Has Productivity Change Counteracted Declining Commodity Producer Prices? 4.8.1. Comparison of the rates of change in agricultural productivity and export unit values In this section, a comparison is made between the rate of change in agricultural productivity and the rate of change in export prices, measured as export unit values. TFP is chosen for the assessment rather than the partial measure of labour productivity because it provides a more comprehensive estimate. Export unit values are selected rather than producer prices because the data on export unit values are more reliable and these border prices are closer to the true economic values than producer prices. As mentioned above, border prices are imperfectly transmitted to producers, and producer prices are distorted by domestic policies that vary over time and between countries. As outlined above, the world export unit value index is used as a deﬂator to ensure constant prices. Given our particular interest in producer welfare, the comparison between rates of TFP and export unit values could be couched in terms of an approximation to the single factoral terms of trade for producers supplying agricultural export markets. Interpreted in this way, it is assumed ﬁrst that the world export unit value index is a good approximation of the import price index for each country. The latter is available for very few countries for the full study period and so could not be used. Second, it is assumed that either the rate of change in TFP in productive activities beyond the farm to the point of export for the selected commodities is the same as that for productive activities on the farm (unlikely) or (much more likely) those ﬁrms undertaking post-farm activities are able to capture the beneﬁts of any differential between the higher TFP rate of change that they achieve and the rate of TFP change achieved by agricultural producers. In section 8.2, we explicitly estimate the single factoral terms of trade at the producer level using an index of producer prices and general consumer prices for each country. These estimates provide a useful check on the consistency of results with those we obtain in this section using country-level export unit

113

The Issue of Declining Commodity Prices values and the world export unit value index. Both indices have their advantages and disadvantages. In particular, export unit values are likely to be more accurate for the selected commodities under study but include price changes between the farm gate and the point of export, the rates of change for which might be different from those facing producers. 4.8.1.1. AGGREGATE CHANGES IN AGRICULTURAL PRODUCTIVITY AND EXPORT UNIT VALUES The crudest way to assess the revenue-enhancing or revenue-reducing effects of changes in commodity outputs and prices is to compare the annual mean rates of change in productivity and real prices over the study period. These estimates are presented in Table 4.8 for all countries, Commonwealth countries, African countries and African Commonwealth countries in the sample. As mentioned above, estimates of productivity change are focused solely on TFP and export unit value indices. Net barter terms of trade for primary commodities have been traditionally measured as the price index of primary products relative to the price index of exports of manufactured products from developed countries (Duncan 1994, p. 56). The United Nations export unit value index for manufactures has typically been used as the denominator in this index. A problem with this denominator, pinpointed by Lipsey (1994) and referred to by Duncan (1994, p. 56), is that its growth can be over-estimated by up to 1 per cent per annum because of failure to account for quality improvements in manufactures exported to developing countries and a lower rate of price increase in manufactured exports to developing countries than to all countries. (The latter is a trivial proportion of the discrepancy.) Lipsey (1994, p. 1) conceded, however, that ‘no conceivable estimate of bias in measures of manufactured goods prices would reverse the picture of declining relative primary product prices during the 1980s’. To limit the extent of these biases, export unit values are deﬂated by the world export unit value index compiled by IMF (2004), which covers all exported products and not just manufactures. The contrast between the aggregate rates of change in TFP and export unit values is stark, with low rates of advance in TFP dwarfed by substantial rates of decline in export unit values. Even allowing for variations in the rates of decline in the two commodity groups in the different country groups, there is Table 4.8. Aggregate Rates of Change in TFP and Export Prices Country group All countries Commonwealth countries African countries African Commonwealth countries

114

Rate of change in TFP (%)

Rate of change in export unit value index (%)

þ0.30 þ0.12 þ0.16 þ0.34

3.17 2.76 3.32 3.03

Analysis of Movements in Productivity and Prices still a gulf between price declines and TFP growth rates. The smallest price decline of 1.85 per cent per annum for ﬁeld crops in the Commonwealth African countries is more than ﬁve times the highest TFP growth rate, also for Commonwealth African countries. The low rate of TFP growth in African countries—0.16 per cent per annum, or only 0.09 per cent if South Africa is excluded—is consistent with the frequently expressed concern about Africa’s lack of success in achieving its own brand of Green Revolution technological progress. Even the modest growth rate for African Commonwealth countries of 0.34 per cent per annum is a bit misleading because exclusion of the southern countries of South Africa, Lesotho, Swaziland, and Namibia from the sample results in a mean decline in TFP of 0.04 per cent per annum. Aggregate ﬁgures, then, can hide more than they reveal. It is therefore necessary to examine rates of price and TFP change in individual countries. As indicated above, rates of change in TFP varied widely over the study period and a few countries did manage to achieve growth rates comparable to the price decline they experienced. 4.8.1.2. COUNTRIES IN WHICH THE RATE OF AGRICULTURAL PRODUCTIVITY GROWTH HAS OUTWEIGHED THE RATE OF DECLINE IN EXPORT UNIT VALUES Sixty-seven of the sampled countries produced at least one of the selected commodities and had a full set of data on export unit values for the period from 1970 to 2002. There were seven countries for which the productivity growth rate in agriculture equalled or exceeded the rate of decline in the producer price index of selected tropical commodities (see Table 4.9). Two of these countries achieved low levels of TFP growth (Jamaica 0.38 per cent and Mauritius 0.21 per cent per annum) but managed to record small positive percentages for commodity price change. Figure 4.16 shows the situation for Jamaica. This commodity price trend was made possible by guaranteed prices in secure markets for their sugar exports originally under the Lome´ agreements and more recently the Cotonou agreement for ACP countries. Barbados also beneﬁted from this market access but suffered signiﬁcant TFP decline. A few other countries, most notably Fiji and Trinidad and Tobago, were in a similar position in that they suffered only a small commodity price decline but, like Barbados, experienced a substantial decline in TFP. Price and TFP trends in Fiji are presented in Figure 4.17. Bangladesh, Gabon, Solomon Islands, and Venezuela achieved low rates of price increase that were augmented by relatively high TFP growth rates of 3.34 per cent, 2.02 per cent, 1.00 per cent and 2.04 per cent per annum, respectively. While the strong performance by Bangladesh appears promising, much of the TFP growth has come from rice production while only very small values

115

The Issue of Declining Commodity Prices Table 4.9. Comparison of Rates of Change in TFP and Selected Commodity Prices Countries with a rate of TFP growth greater than the rate of commodity price decline: Bangladesh Laos Solomon Islands Gabon Mauritius Venezuela Jamaica Countries with a TFP growth rate at least one-half the rate of commodity price decline: Benin Costa Rica South Africa Bolivia Cuba Swaziland Central African Republic Nigeria Countries with a TFP growth rate less than one-half the rate of commodity price decline: Argentina Indonesia Rwanda ˆ te d’Ivoire Co Kenya Sri Lanka China Madagascar Syria Colombia Malawi Tanzania Egypt Malaysia Turkey Guatemala Mali Uganda Haiti Mexico Uruguay India Papua New Guinea Zimbabwe Countries with a negative rate of TFP change: Angola El Salvador Barbados Fiji Brazil Ghana Burkina Faso Guinea Burundi Honduras Cameroon Liberia Chad Mozambique Congo, Democratic Republic Nicaragua Dominican Republic Niger Ecuador

Paraguay Peru Senegal Sierra Leone Sudan Togo Trinidad and Tobago Vanuatu Zambia

Jamaica 2.5 TFP Price index

Index

2.0

Linear (Price index)

1.5

1.0

0.5

19 7 19 0 7 19 1 7 19 2 7 19 3 7 19 4 7 19 5 7 19 6 7 19 7 7 19 8 7 19 9 8 19 0 8 19 1 8 19 2 8 19 3 8 19 4 8 19 5 8 19 6 8 19 7 8 19 8 8 19 9 9 19 0 9 19 1 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 9 19 7 9 19 8 9 20 9 0 20 0 0 20 1 02

0.0

Figure 4.16. Trends in Export Unit Values and TFP in Jamaica, 1970 to 2002

116

Analysis of Movements in Productivity and Prices Fiji 3.0 TFP Price index Linear (Price index)

2.5

Index

2.0

1.5

1.0

0.5

19 7 19 0 7 19 1 7 19 2 7 19 3 7 19 4 7 19 5 7 19 6 7 19 7 7 19 8 7 19 9 8 19 0 8 19 1 8 19 2 8 19 3 8 19 4 8 19 5 8 19 6 8 19 7 8 19 8 8 19 9 9 19 0 9 19 1 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 9 19 7 9 19 8 9 20 9 0 20 0 0 20 1 02

0.0

Figure 4.17. Trends in Export Unit Values and TFP in Fiji, 1970 to 2002

Solomon Islands 5.0 4.5

TFP

4.0

Price index

3.5

Linear (Price index)

Index

3.0 2.5 2.0 1.5 1.0 0.5

19 7 19 0 7 19 1 7 19 2 7 19 3 7 19 4 7 19 5 7 19 6 7 19 7 7 19 8 7 19 9 8 19 0 8 19 1 8 19 2 8 19 3 8 19 4 8 19 5 8 19 6 8 19 7 8 19 8 8 19 9 9 19 0 9 19 1 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 9 19 7 9 19 8 9 20 9 0 20 0 0 20 1 02

0.0

Figure 4.18. Trends in Export Unit Values and TFP in Solomon Islands, 1970 to 2002

are earned from exports of the selected commodities relative to total agricultural production. Even though it exported small amounts of most of the selected commodities at different times, Venezuela is in a similar position to Bangladesh in that exports of selected commodities contribute only a small proportion of its total agricultural value of output. Gabon and Solomon Islands (see Figure 4.18) depend quite strongly on selected commodity exports

117

The Issue of Declining Commodity Prices and beneﬁted from high cocoa and palm oil prices towards the end of the study period. But exports of cocoa beans from Gabon declined in the ﬁnal ﬁve years of the period. Finally, the rate of growth of TFP in Lao PDR of 2.71 per cent outweighed the commodity price decline of 1.54 per cent. As in Bangladesh, TFP increase came mainly from improved technologies in rice production (from a low base in 1970) and selected commodity exports are a small proportion of total agricultural production, comprising some coffee and a tiny bit of cotton. 4.8.1.3. COUNTRIES IN WHICH THE AGRICULTURAL PRODUCTIVITY GROWTH RATE HAS NOT MATCHED THE RATE OF DECLINE IN EXPORT UNIT VALUES Table 4.9 contains a list of 61 countries in the sample for which the rate of TFP growth in agriculture was less than the rate of decline in the producer price index of selected tropical commodities. Of these 61 countries, 28 had negative rates of change in TFP that augmented commodity price decline. Of the remaining 31 countries, 24 countries had rates of productivity growth that were less than one-half the rate of decline in commodity prices. In many cases, the choice of study period is important. For example, Ghana is one of the countries with a negative rate of change in TFP over the full study period (see Figure 4.19). Yet, as mentioned above, the country experienced a solid rate of TFP growth in the ﬁnal two decades of the period, at a rate of 3.53 per cent per annum, while the rate of commodity price decline was slightly lower, at 3.19 per cent per annum.

Ghana 4.0 TFP 3.5

Price index Linear (Price index)

3.0

Index

2.5 2.0 1.5 1.0 0.5

19 7 19 0 7 19 1 7 19 2 7 19 3 7 19 4 7 19 5 76 19 7 19 7 7 19 8 7 19 9 8 19 0 8 19 1 8 19 2 8 19 3 8 19 4 8 19 5 8 19 6 8 19 7 8 19 8 8 19 9 9 19 0 91 19 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 9 19 7 9 19 8 9 20 9 0 20 0 0 20 1 02

0.0

Figure 4.19. Trends in Export Unit Values and TFP in Ghana, 1970 to 2002

118

Analysis of Movements in Productivity and Prices Paradoxically, producers in Ghana suffered a decline in TFP during the 1970s when export unit values were at their zenith and substantial income gains were to be had from increasing output, yet managed to achieve respectable gains in TFP as export prices declined sharply after 1980. The agricultural sector was experiencing a revival during this period, brought about by a vastly improved economic climate, and the upward productivity trend did not cease until 2001. Economic policy shifts largely explain these trends, as the policy direction in the 1970s did not favour the traded goods sector. The beneﬁcial impact of economic liberalization in the 1990 together with reduced taxes on the traded goods sector, aided agricultural development and helps to explain the continued upward trend in TFP until recently. Many other countries achieved modest increases in TFP that failed to match the rate of decline in export unit values. Examples of particular interest in this study are presented in Figure 4.20. In all of these countries, as well as in Ghana as described above, economic reform policies encouraged by the International Monetary Fund and World Bank in the late 1980s and 1990s provided incentives to expand activity in the traded goods sector and encouraged agricultural exports. Increases occurred in productivity from the early 1990s after decades of indifferent performance. In some countries, such as Nigeria and to a lesser extent India, the TFP growth rate almost matches the export price decline. The rate of productivity growth was minor in others, exempliﬁed by Kenya, Sri Lanka, and Tanzania in Figure 4.20, but in contrast to the situation in India and Nigeria productivity gains did little to overcome a steep decline in export unit values over the whole study period. The Kenyan agricultural sector underwent a long process of structural reform from the early 1980s. Reform ‘proceeded slowly until 1993, when pace gradually increased’ (Nyangito and Karugia 2002, p. 151). It had mixed effects on the productive capacity of Kenyan agriculture, as Nyangito and Karugia make clear, but a 20 per cent increase in TFP was achieved between 1993 and 2002 (Figure 4.20). A major factor in this increase was expansion in the horticultural sector after 1995, which followed a period when this sector had failed to live up to its ‘initial promise’ (Roberts and Fagerna¨s 2004, p. 39). India provides an interesting case study given that the Green Revolution technological advances had virtually run their course by the mid 1980s. Concerns were being expressed that a turndown in agricultural productivity was imminent as a result of environmental degradation. Yet the implementation of liberalization policies brought about further impetus for productivity gains that lasted until 2001. As in Ghana, Jamaica, and many other countries, however, these gains turned to decline in the closing years of the study period. This trend change is clearly seen in Figure 4.20 where TFP peaked prior to 2002. In some respects, Malaysia shared the experiences of the countries featuring in Figure 4.20 in that TFP growth was ﬂat in the face of steeply declining export unit values (see Figure 4.21). In contrast to these countries, though, TFP

119

The Issue of Declining Commodity Prices Cote d'Ivoire

6.0

5.0

Nigeria

TFP

4.5

TFP 5.0

Price index

4.0

Price index

Linear (Price index)

Linear (Price index)

3.5

4.0

Index

Index

3.0 2.5 2.0

3.0

2.0

1.5 1.0

1.0

0.5 0.0

2.5

19 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 2099 2000 2001 02

19 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 2099 2000 2001 02

0.0

India

4.0

Papua New Guinea TFP

TFP

3.5

Linear (Price index)

3.0

Linear (Price index)

2.5

Index

1.5

Index

Price index

Price index

2.0

1.0

2.0 1.5 1.0

0.5

0.5 0.0

0.0 19 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 2099 2000 2001 02

19 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 2099 2000 2001 02

6.0

Kenya

3.5

Sri Lanka TFP

TFP 5.0

3.0

Price index

Price index Linear (Price index)

Linear (Price index)

2.5

Index

Index

4.0

3.0

2.0 1.5

2.0 1.0 1.0

0.5

0.0

3.0

19 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 2099 2000 2001 02

19 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 2099 2000 2001 02

0.0

Malawi

4.5 TFP

2.5

4.0

Price index

Index

Index

Linear (Price index)

3.0

2.0

1.0

TFP Price index

3.5

Linear (Price index)

1.5

Tanzania

2.5 2.0 1.5 1.0

19 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 2099 2000 2001 02

0.0

0.5 0.0 19 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 2099 2000 2001 02

0.5

Figure 4.20. Selected Countries with Lower Rate of TFP Growth to Rate of Decline in Export Unit Value, 1970 to 2002

actually declined throughout most of the 1990s then increased slightly from 1998 to 2002. The 1990s were a period of rapid industrialization with an exodus of many people from rural areas to the cities. There was considerable substitution of machines for labour that is normally associated with technological progress and rising TFP, yet this process did not appear to have become established until 1998. Six examples of the many countries that experienced average declines in both export unit values and TFP are presented in Figure 4.22. Two of these countries (Cameroon and Mozambique) show the same upward trend in TFP

120

Index 2.5

1.0

6.0

5.0

Index

Index 2.5

2.0 1.5

Index

Price index

3.0

1.5

0.0

2.0

3.0

1.0

0.0 19 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 2099 2000 2001 02

19 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 2099 2000 2001 02

3.5

19 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 2099 2000 2001 02

Index 4.0

Index

19 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 2099 2000 2001 02

4.5

TFP

Linear (Price index)

1.5

0.0

Sierra Leone

Price index

4.0

19 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 2099 2000 2001 02

19 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 2099 2000 2001 02

19 1970 7 19 1 7 19 2 7 19 3 7 19 4 1975 7 19 6 7 19 7 7 19 8 7 19 9 8 19 0 1981 8 19 2 1983 8 19 4 8 19 5 1986 8 19 7 1988 8 19 9 1990 9 19 1 9 19 2 9 19 3 1994 9 19 5 9 19 6 9 19 7 9 19 8 2099 0 20 0 0 20 1 02

Index

Analysis of Movements in Productivity and Prices

4.5 Malaysia

4.0 TFP Price index Linear (Price index)

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0.0

Figure 4.21. Trends in Export Unit Values and TFP in Malaysia, 1970 to 2002 Cameroon TFP 3.0

Trinidad and Tobago

Linear (Price index) 2.5 TFP

2.0

Mozambique

TFP

7.0

Linear (Price index)

6.0 Price index

Linear (Price index)

1.0 1.0

0.5 0.5

0.0

8.0

7.0

Uganda

Price index

6.0 TFP

Price index

5.0 Linear (Price index)

4.0

0.5

3.0

2.0

1.0

0.0

Vanuatu

5.0

TFP

Price index

Linear (Price index)

4.0

2.0

3.0

2.0

1.0

0.0

Figure 4.22. Selected Countries Experiencing Rates of Decline in Export Unit Values and TFP, 1970 to 2002

121

The Issue of Declining Commodity Prices in the 1990s, peaking late in the decade, which was evident for countries reported in Figure 4.20. The main difference, however, is that Cameroon and Mozambique went through long periods of declining agricultural productivity in the ﬁrst two decades of the study period. Trinidad and Tobago also underwent a major decline in TFP from 1970 until the mid-1980s, and ﬂuctuated around a constant trend thereafter. Uganda has gone through a chequered process with substantial increases in TFP prior to the 1970s that continued throughout that decade at a moderate rate before stagnating in the 1980s and declining in the 1990s. TFP was virtually unchanged in Sierra Leone until 1998 after which it declined quite markedly as political conditions worsened. Small island countries such as Tonga and Vanuatu rely heavily on tree crop exports. Producers in these two countries not only failed to counter declining export prices for these commodities but also experienced declining TFP (see Figure 4.22 for Vanuatu). This trend was due in no small measure to the declining fortunes of the coconut industry that was still a major force at the beginning of the study period in each country. TFP declined annually by 1.27 per cent in Vanuatu and 1.02 per cent in Tonga.

4.8.2. Trends in the single factoral terms of trade for producers of selected commodities In this section, we provide estimates of the single factoral terms of trade for each country under study for the period from 1970 to 2002. As indicated above, it was not possible to get full sets of data for the thirty-two years for any country because of deﬁciencies in either producer price or consumer price data, especially for the period from 1999 to 2002. Nevertheless, we have managed to compile a data set that provides a reasonably comprehensive picture of trends in the single factoral terms of trade for 63 countries. Results show that annual producer price movements do not correspond very closely to those of export unit values of the selected commodities. A 10 per cent increase in the annual export unit value index led to at least a 5 per cent increase in the annual producer price index in only six of the 54 countries for which it was possible to regress the producer price index on the export unit value index. The coefﬁcient on the export unit value index is actually negative in 13 of the countries. The varied inﬂuences of government economic policies, especially those inﬂuencing agricultural export industries, make this ﬁnding unsurprising. Where they were in operation for substantial parts of the study period, agricultural commodity stabilization schemes would have greatly reduced ﬂuctuations in producer prices in many countries. While short-term ﬂuctuations are not correlated, the long-term trends in the two price indices that are of interest in this study are broadly similar.

122

Analysis of Movements in Productivity and Prices In the three sections that follow, we identify countries in which the single factoral terms of trade improved, did not change or deteriorated. Brief explanations are given of the identiﬁed trends, concentrating for the most part on Commonwealth and African countries in which the selected commodities comprise important exports in their agricultural sectors. In all diagrams presented in this section, trend lines are added using a polynomial curve. 4.8.2.1. COUNTRIES IN WHICH THE SINGLE FACTORAL TERMS OF TRADE IMPROVED Table 4.10 contains a list of the ﬁfteen countries in the sample for which the single factoral terms of trade in the selected tropical commodities improved over the study period. Most of the countries in this group rely at least to a moderate extent on the production and export of the selected commodities: Central African Republic (coffee and cotton); Chad (cotton); Colombia (coffee, cotton, palm oil, palm kernel oil and sugar); Guatemala (coffee, cotton, palm oil and sugar); Indonesia (cocoa, coffee, palm oil, lauric oils, rice); Jamaica (cocoa, coffee and Table 4.10. Trends in the Single Factoral Terms of Trade Countries with a positive trend in the single factoral terms of trade: Bangladesh India Central African Republic Indonesia Chad Iran Colombia Jamaica Guatemala Madagascar Countries with no signiﬁcant trend in the single factoral terms of trade: China Mali Costa Rica Mauritania Gabon Mauritius Laos Mozambique Malawi Solomon Islands Countries with a negative trend in the single factoral terms of trade: Barbados Fiji Bolivia Gambia, The Botswana Ghana Burkina Faso Guinea Burundi Guinea-Bissau ˆ te d’Ivoire Co Haiti Cameroon Honduras Dominican Republic Kenya Ecuador Malaysia Egypt Mexico El Salvador Morocco Burkina Faso Guinea Burundi Honduras Cameroon Liberia Chad Mozambique Congo, Democratic Republic Nicaragua Dominican Republic Niger Ecuador

Nepal Nigeria Papua New Guinea South Africa Togo Sri Lanka Swaziland Uruguay Venezuela

Nigeria Paraguay Rwanda Senegal Sierra Leone Sudan Tonga Trinidad and Tobago Vanuatu Zambia Zimbabwe

123

The Issue of Declining Commodity Prices

Nigeria 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 19 7 19 0 7 19 1 7 19 2 7 19 3 7 19 4 7 19 5 7 19 6 7 19 7 7 19 8 7 19 9 8 19 0 8 19 1 8 19 2 8 19 3 8 19 4 8 19 5 8 19 6 8 19 7 8 19 8 8 19 9 9 19 0 9 19 1 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 9 19 7 9 19 8 9 20 9 0 20 0 0 20 1 02

Single factoral terms of trade (1990=100)

sugar); Madagascar (cocoa, coffee, rice and sugar); Nigeria (cocoa, cotton, palm oil, palm kernel oil); Papua New Guinea (coffee, cocoa, palm oil and lauric oils); and Togo (cotton, cocoa and coffee). The trend in productivity and prices in Jamaica is discussed above, with sluggish TFP growth but an increase in domestic prices particularly in the late 1980s and early 1990s. Two of the countries in the group, Central African Republic (selected commodity exports of coffee and cotton) and Nigeria (the main selected commodity exports being cocoa, cotton and palm oil), are shown in Table 4.9 to have TFP growth rates approaching the rate of decline in export unit values. In Nigeria, domestic commodity prices increased at a much faster rate than inﬂation in the ﬁrst half of the 1990s, a trend vividly illustrated in Figure 4.23. The situation changed from 1995 to 1998 when the consumer price index increased substantially while domestic commodity prices stagnated. The polynomial trend line shown in Figure 4.23 demonstrates this effect, tipping the index downwards towards the end of the study period. A full data set was only available for a relatively short period for the Central African Republic (1980–1998) (see Figure 4.24) and Chad (1983–98). During this period, the agricultural sector in Chad performed much better in terms of TFP growth than for the whole study period, and also experienced a slight increase in domestic prices of the selected commodities. The Central African Republic had a relatively high rate of TFP growth over the whole study period, albeit from a very low base, and domestic prices increased relative to the consumer price index. The picture is probably less rosy for the full period because TFP declined by 10 per cent between 1997 and 2002.

Figure 4.23. Trends in the Single Factoral Terms of Trade in Nigeria, 1970 to 2002

124

Analysis of Movements in Productivity and Prices Central African Republic Single factoral terms of trade (1990=100)

1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0

1980198119821983198419851986198719881989199019911992199319941995199619971998

Figure 4.24. Trends in the Single Factoral Terms of Trade in the Central African Republic, 1970 to 1998

In Indonesia, domestic prices were maintained in real terms until 1998, while TFP growth took place at a modest rate. It is of concern that domestic prices declined substantially in real terms from 1998 to 2001 during the period of ﬁnancial crisis (the last year for which domestic prices are available). This trend was not of sufﬁcient length to have a marked deleterious effect on the factoral terms of trade, but is likely to have had an effect for a data set extending beyond 2001. Producers of the selected commodities in Colombia, Guatemala, Madagascar and Papua New Guinea achieved low to moderate rates of TFP growth, but also beneﬁted from domestic prices that rose much more quickly than either the consumer price index or the index of export unit values at various stages of the 1990s. Domestic prices increased particularly fast in relation to the consumer price index in that decade. They increased at twice the rate of consumer prices throughout the 1990s in Colombia, and increased by 116 per cent in 1995 in Guatemala and by 83 per cent in 1994 in Madagascar while the consumer price indices increased by only 8 per cent and 39 per cent respectively. In the case of Papua New Guinea, devaluation of the local currency in the 1990s helped maintain the domestic price of its exports while the consumer price index did not increase to the same extent (see Figure 4.25). Towards the end of the period for which data are available (1998), relative prices began to change and since 1997 levels of domestic commodity prices have fallen well behind general price levels. In this respect, the overall favourable trend in the factoral terms of trade to 1998 shown here paints too optimistic a picture.

125

The Issue of Declining Commodity Prices Papua New Guinea Single factoral terms of trade (1990=100)

1.4 1.2 1.0 0.8 0.6 0.4 0.2

19 7 19 0 7 19 1 7 19 2 7 19 3 7 19 4 7 19 5 7 19 6 77 19 7 19 8 7 19 9 8 19 0 8 19 1 8 19 2 8 19 3 8 19 4 8 19 5 8 19 6 8 19 7 8 19 8 8 19 9 9 19 0 9 19 1 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 9 19 7 98

0.0

Figure 4.25. Trends in the Single Factoral Terms of Trade in Papua New Guinea, 1970 to 1998

4.8.2.2. COUNTRIES IN WHICH THE SINGLE FACTORAL TERMS OF TRADE DID NOT CHANGE SIGNIFICANTLY Fourteen countries experienced rates of TFP growth that were not signiﬁcantly different from the rate of change in the domestic prices of the selected commodities. Eight of these countries have signiﬁcant exports of the selected commodities: Costa Rica (coffee, cocoa, palm oil, palm kernel oil, sugar); Gabon (cocoa, coffee, palm oil); Malawi (coffee, cotton, sugar); Mali (cotton); Mauritius (sugar); Mozambique (copra, coconut oil, cotton, sugar); Solomon Islands (cocoa, lauric oils, palm oil); and Sri Lanka (lauric oils, rice). A feature of all but two of these countries is an increase in the domestic commodity price index relative to the consumer price index in the early to mid-1990s followed by a decline later in the decade. Figure 4.26 dramatically demonstrates these effects on the single factoral terms of trade for Costa Rica. Following an extended period in which it was static from the late 1970s until 1990, the single factoral terms of trade index rose sharply to 1994 and then deteriorated even more sharply until 1998. This set of circumstances was typically associated with devaluation of the local currency, usually as part of an economic reform program, that fed through to higher commodity export prices in local currency terms. However, these gains were soon eroded by a rise in the consumer price index in response to higher import prices in local currency terms. The decline in the ratio of commodity prices to the consumer price index was slighter in Mauritius, as indicated in Figure 4.27, but sugar producers in this country and other ACP countries face an uncertain future in the wake of any loss of their privileged

126

Analysis of Movements in Productivity and Prices Costa Rica

1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 19 7 19 0 7 19 1 7 19 2 7 19 3 7 19 4 75 19 7 19 6 7 19 7 7 19 8 7 19 9 8 19 0 81 19 8 19 2 8 19 3 8 19 4 8 19 5 8 19 6 87 19 8 19 8 8 19 9 9 19 0 9 19 1 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 9 19 7 9 19 8 9 20 9 0 20 0 01

Single factoral terms of trade (1990=100)

2.0

Figure 4.26. Trends in the Single Factoral Terms of Trade in Costa Rica, 1970 to 2002

Single factoral terms of trade (1990=100)

Mauritius 3.5 3.0 2.5 2.0 1.5 1.0 0.5

19 7 19 0 7 19 1 7 19 2 73 19 7 19 4 7 19 5 7 19 6 7 19 7 7 19 8 7 19 9 8 19 0 8 19 1 8 19 2 8 19 3 8 19 4 8 19 5 8 19 6 8 19 7 8 19 8 8 19 9 9 19 0 9 19 1 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 9 19 7 98

0.0

Figure 4.27. Trends in the Single Factoral Terms of Trade in Mauritius, 1970 to 1998

access to European sugar markets. The key role of commodity export prices is demonstrated in Figure 4.27, with a spectacular price-induced improvement occurring during the commodity boom of the mid-1970s followed by an equally spectacular decline in the late 1970s. As mentioned above, most of the productivity gains that fed through to an improvement in the single factoral terms of trade in Solomon Islands were associated with development of the oil palm industry, and rehabilitation of the coconut industry and smallholder cocoa production in the 1970s. The single factoral terms of trade actually declined slightly in the 1980s and 1990s,

127

The Issue of Declining Commodity Prices

Single factoral terms of trade (1990=100)

Solomon Islands 2.5

2.0

1.5

1.0

0.5

19 7 19 0 7 19 1 7 19 2 7 19 3 7 19 4 75 19 7 19 6 7 19 7 7 19 8 7 19 9 8 19 0 8 19 1 8 19 2 8 19 3 8 19 4 8 19 5 8 19 6 8 19 7 8 19 8 8 19 9 9 19 0 9 19 1 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 9 19 7 98

0.0

Figure 4.28. Trends in the Single Factoral Terms of Trade in Solomon Islands, 1970 to 1998

Sri Lanka Single factoral terms of trade (1990=100)

1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2

19 7 19 0 7 19 1 7 19 2 7 19 3 7 19 4 7 19 5 7 19 6 7 19 7 7 19 8 7 19 9 8 19 0 8 19 1 8 19 2 8 19 3 8 19 4 8 19 5 8 19 6 8 19 7 8 19 8 8 19 9 9 19 0 9 19 1 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 9 19 7 9 19 8 9 20 9 0 20 0 01

0.0

Figure 4.29. Trends in the Single Factoral Terms of Trade in Sri Lanka, 1970 to 2001

as indicated by the polynomial trend line in Figure 4.28. However, increases occurred in the unit values of commodity exports late in the study period, from 1998, and they fed through to domestic prices (the incipient effects of which are shown in Figure 4.27). But the future proﬁtability of the oil palm, lauric oil and cocoa export industries has been clouded by recent civil unrest. Unlike the countries discussed above, Sri Lanka did not receive a boost in the single factoral terms of trade from an increase in domestic commodity prices relative to the consumer price index in the 1990s. In fact, this price ratio fell by one-third between 1990 and 2001, the last year for which data are available.

128

Analysis of Movements in Productivity and Prices The reason why the single factoral terms of trade did not signiﬁcantly deteriorate overall lies in the events in the ﬁrst few years of the study period. The index rose from 0.58 in 1970 to 1.42 in 1981 and deteriorated thereafter, reaching 0.75 in 2001 (Figure 4.29). If only the two decades from 1981 to 2001 were examined for trend, the single factoral terms of trade would have shown an annual decline of 2.3 per cent. 4.8.2.3. COUNTRIES IN WHICH THE SINGLE FACTORAL TERMS OF TRADE DETERIORATED Table 4.10 also contains a list of countries in the sample for which the single factoral terms of trade in selected tropical commodities deteriorated over the study period. Trends in the single factoral terms of trade in Kenya (a slight decline of 1.4 per cent per annum) are typical of many countries suffering deterioration (Figure 4.30). Moderate gains in TFP have been more than offset by a decline in the net barter terms of trade to producers of the selected commodities (principally coffee but also erratic exports of palm oil, sugar, lauric oils and cotton in the case of Kenya). A complicating factor here is that most of the gains in TFP have probably been in horticultural industries while other agricultural industries such as coffee have suffered declining productivity. Shikwati and Okonski (2005, p. 1) recently reported that: The decline in coffee earnings has contributed to low productivity in what was once Kenya’s ‘black gold’. Coffee berry diseases, leaf rust, leached soils, high input and marketing costs have made Kenyan farmers invest less in this sector.

A contrasting picture is painted for Sierra Leone where descent into civil war has led to a massive decline in the single factoral terms of trade for producers of

Single factoral terms of trade (1990=100)

Kenya 3.5 3.0 2.5 2.0 1.5 1.0 0.5

19 7 19 0 71 19 7 19 2 73 19 7 19 4 75 19 7 19 6 77 19 7 19 8 7 19 9 8 19 0 8 19 1 8 19 2 83 19 8 19 4 85 19 8 19 6 87 19 8 19 8 89 19 9 19 0 91 19 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 97 19 98

0.0

Figure 4.30. Trends in the Single Factoral Terms of Trade in Kenya, 1970 to 1998

129

The Issue of Declining Commodity Prices Sierra Leone Single factoral terms of trade (1990=100)

3.0 2.5 2.0 1.5 1.0 0.5

19 7 19 0 71 19 7 19 2 73 19 7 19 4 7 19 5 7 19 6 77 19 7 19 8 79 19 8 19 0 81 19 8 19 2 83 19 8 19 4 85 19 8 19 6 8 19 7 8 19 8 89 19 9 19 0 91 19 9 19 2 93 19 9 19 4 95 19 9 19 6 97 19 98

0.0

Figure 4.31. Trends in the Single Factoral Terms of Trade in Sierra Leone, 1970 to 1998

the major export crops of cocoa and coffee (Figure 4.31). The rapid decline in the index was caused by the interaction of a gradually declining TFP and rapidly declining net barter terms of trade facing producers during the 1990s. Finally, the trend shown above in the single factoral terms of trade index of a sugar-exporting country, Mauritius, contrasts with a number of other sugarexporting countries in which producers have faced secularly declining single factor terms of trade despite a secure export market. Trinidad and Tobago is one example, shown in Figure 4.32, where declining TFP during the 1990s due to

Trinidad and Tobago Single factoral terms of trade (199=100)

4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 19 7 19 0 71 19 7 19 2 73 19 7 19 4 75 19 7 19 6 7 19 7 7 19 8 7 19 9 80 19 8 19 1 8 19 2 8 19 3 84 19 8 19 5 86 19 8 19 7 8 19 8 8 19 9 9 19 0 91 19 9 19 2 93 19 9 19 4 95 19 9 19 6 97 19 98

0.0

Figure 4.32. Trends in the Single Factoral Terms of Trade in Trinidad and Tobago, 1970 to 1998

130

Analysis of Movements in Productivity and Prices Appendix 4.1: Values of selected tropical commodities as a proportion of export values and total agricultural product

Country

Export values of selected tropical commodities as a: Percentage Percentage of total of total agricultural exports FOB product Country

Uganda Burundi Rwanda El Salvador Ghana ˆ te d’Ivoire Co Guatemala Honduras

79.24% 77.72% 66.27% 41.53% 40.15% 34.88% 26.75% 21.38%

81.50% 81.48% 70.93% 79.93% 87.23% 62.96% 40.44% 29.21%

Nicaragua Colombia Costa Rica Kenya Cameroon Papua New Guinea Madagascar Dominican Republic

20.48% 20.17% 18.91% 17.84% 15.40% 14.55% 13.21% 12.63%

28.08% 59.85% 29.94% 28.29% 59.11% 77.82% 23.91% 24.33%

Nepal Sierra Leone Central African Rep. Congo, Republic Of Togo Ecuador Haiti Malaysia Benin Brazil Peru Indonesia Malawi Bolivia Zimbabwe Guinea Jamaica

10.10% 9.76% 8.44% 8.17% 6.71% 6.56% 6.47% 6.24% 5.23% 3.93% 3.05% 2.81% 2.69% 1.90% 1.35% 1.09% 1.06%

45.59% 80.10% 28.01% 81.54% 28.83% 22.64% 56.74% 41.24% 14.71% 14.09% 36.81% 26.92% 2.90% 8.05% 3.35% 26.67% 5.59%

Nigeria Paraguay Mexico India Swaziland Sudan Zambia Trinidad And Tobago Mali Sri Lanka Venezuela Senegal Gabon Niger Angola China, People’s Republic Uruguay Morocco Botswana South Africa Argentina Bangladesh Burkina Faso Chad Chile Gambia, The Guinea-Bissau Mauritania Mozambique Myanmar Namibia Tanzania

Export values of selected tropical commodities as a: Percentage of total exports FOB

Percentage of total agricultural product

1.01% 0.99% 0.82% 0.69% 0.57% 0.35% 0.33%

59.99% 2.57% 11.36% 4.09% 0.91% 0.21% 17.15%

0.30% 0.24% 0.23% 0.21% 0.16% 0.12% 0.12% 0.12%

5.29% 0.32% 0.58% 11.86% 0.67% 91.34% 0.97% 82.21%

0.07% 0.06% 0.01% 0.01% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

0.35% 0.14% 0.07% 0.19% 0.08% 0.00% 0.01% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%

131

The Issue of Declining Commodity Prices Appendix 4.2: Annual rates of change in export quantities of selected commodities Variable All countries Total commodities: Intercept Trend R square Tree crops: Intercept Trend R square Field crops: Intercept Trend

Coefﬁcient

Standard error

t-statistic

p-value

12.332 0.018

0.039 0.001

315.380 12.687

<0.001 <0.001

Standard error

0.077

333.664 22.377

<0.001 0.001

Standard error

0.067

183.850 3.858

<0.001 0.001

Standard error

0.129

213.647 19.784

<0.001 <0.001

Standard error

0.091

250.729 27.850

<0.001 <0.001

Standard error

0.072

86.338 6.519

<0.001 <0.001

Standard error

0.209

196.867 5.452

<0.001 <0.001

Standard error

0.103

193.508 3.831

<0.001 <0.001

Standard error

0.100

126.923 5.635

<0.001 <0.001

Standard error

0.146

151.159 0.912

<0.001 0.369

Standard error

0.128

141.096 0.597

<0.001 0.555

Standard error

0.130

104.789 2.155

<0.001 0.039

Standard error

0.168

0.839 11.239 0.027 0.942 12.035 0.009

R square Commonwealth countries Total commodities: Intercept Trend

0.324

R square Tree crops: Intercept Trend

0.927

R square Field crops: Intercept Trend

0.962

R square African countries Total commodities: Intercept Trend

0.578

9.791 0.033

9.160 0.037

9.113 0.025

10.295 0.010

R square Tree crops: Intercept Trend

0.489

R square Field crops: Intercept Trend

0.321

9.803 0.007

9.351 0.015

R square 0.506 African Commonwealth countries Total commodities: Intercept 9.805 Trend 0.002 R square Tree crops: Intercept Trend

0.065 0.002

0.046 0.002

0.037 0.001

0.106 0.004

0.052 0.002

0.051 0.002

0.074 0.003

0.065 0.002

0.026 9.288 0.001

R square Field crops: Intercept Trend

0.011

R square

0.130

132

0.034 0.001

8.907 0.007

0.066 0.002

0.085 0.003

Analysis of Movements in Productivity and Prices Appendix 4.3: Annual rates of change in TFP, labour productivity and export unit values of selected commodities

Country Algeria Angola Argentina Bangladesh Barbados Benin Bolivia Botswana Brazil Burkina Faso Burundi ˆ te d’Ivoire Co Cambodia Cameroon Central African Republic Chad Chile China Colombia Congo, Democratic Republic of Costa Rica Cuba Dominican Republic Ecuador Egypt El Salvador Fiji Islands Gabon Gambia, The Ghana Guatemala Guinea Guinea-Bissau Haiti Honduras India Indonesia Iran Jamaica Kenya Laos Lesotho Liberia Madagascar Malawi Malaysia Mali Mauritania Mauritius Mexico

Annual rate of change in TFP (%)

Annual rate of change in labour productivity (%)

Annual rate of change in export unit value index (%)

0.66 0.49 0.10 3.34 2.18 1.49 1.31 0.28 0.71 1.90 0.69 0.68 0.51 0.92

0.02 1.49 3.28 0.86 0.47 3.64 1.84 0.37 3.83 1.86 0.69 1.68 1.05 1.61

n.a. 5.62 2.47 0.43 0.44 2.88 1.78 n.a. 0.63 1.88 3.28 3.98 n.a. 4.15

2.36 0.69 4.27 0.36 1.32

0.49 0.73 2.01 2.26 1.59

3.46 0.50 n.a. 2.64 3.05

4.44 2.89 2.90

0.88 2.53 1.39

5.26 3.01 5.58

0.22 1.04 0.16 0.62 2.48 2.01 1.68 0.24 1.04 0.36 2.12 0.37 1.20 1.44 1.57 0.71 0.38 1.16 2.71 0.19 0.03 0.37 0.50 0.33 1.37 0.27 0.21 0.39

0.92 0.97 2.66 0.04 1.35 3.34 2.49 0.28 1.18 0.16 1.01 0.52 0.57 1.18 1.86 3.35 0.33 0.14 1.54 0.02 0.53 0.90 1.02 4.27 1.79 1.30 1.30 1.33

1.72 2.97 2.26 2.93 1.17 0.19 n.a. 3.22 3.50 5.72 n.a. 2.35 2.46 3.16 3.47 n.a. 0.42 2.52 1.54 n.a. 5.24 3.71 1.34 2.96 2.75 n.a. 0.18 2.37 (continued )

133

The Issue of Declining Commodity Prices Appendix 4.3: (Continued)

Country Mongolia Morocco Mozambique Namibia Nepal Nicaragua Niger Nigeria Papua New Guinea Paraguay Peru Rwanda Saudi Arabia Senegal Sierra Leone Solomon Islands South Africa Sri Lanka Sudan Swaziland Syria Tanzania Togo Tonga Trinidad and Tobago Tunisia Turkey Uganda Uruguay Vanuatu Venezuela Zambia Zimbabwe

Annual rate of change in TFP (%)

Annual rate of change in labour productivity (%)

Annual rate of change in export unit value index (%)

1.03 0.21 1.47 2.19 1.99 1.29 1.28 1.97 0.02 0.42 0.27 0.32 0.07 1.25 1.04 1.00 2.75 0.79 0.08 2.18 1.46 0.35 0.17 1.02 2.53 1.28 0.66 0.57 0.84 1.27 2.04 1.21 0.58

0.29 1.06 1.13 1.19 1.16 0.23 0.25 2.77 0.61 2.37 0.98 0.43 6.72 1.10 0.72 0.20 3.28 0.08 0.18 1.55 2.94 0.36 0.53 0.03 1.41 1.93 1.72 0.76 2.13 0.87 2.32 0.31 0.29

n.a. n.a. 2.57 n.a. n.a. 2.23 4.98 3.64 2.68 2.14 2.03 3.25 n.a. 1.74 4.10 0.28 3.24 1.66 3.71 2.44 3.23 2.60 3.39 n.a. 0.21 n.a. 4.86 4.84 2.35 2.56 0.85 0.71 1.43

falling sugar and cocoa yields marginally offset the advantages of a guaranteed export price for sugar and considerable domestic price support provided by the government in the ﬁnal years of the study period (Knapp 2004, p. 17).

4.8.3. Summary The estimates of the single factoral terms of trade outlined above reveal that more than twice as many countries experienced deteriorating trends as improving trends over the different periods for which data are available. Full data sets were available only until 1998 for almost all countries, which meant that the results do not take into account what occurred in the ﬁnal ﬁve years of the study

134

Analysis of Movements in Productivity and Prices period. A review of the known trends in TFP for this sub-period suggests that results for the full study period would have been less promising than those reported here. Nevertheless, trend results are broadly in line with those obtained comparing TFP changes with export unit values deﬂated by the world export unit value index that show a deterioration in the returns to factors engaged in the production of commodity exports.

135

5 Commodity Value Chains Compression—Coffee, Cocoa, and Sugar Jaya Choraria

Although there have been numerous quantitative studies providing evidence to illustrate the extent of the problems of commodity price volatility and the long run secular decline in the price of commodities vis-a`-vis manufactured goods, there is a lack of quantitative analysis of the evolution of the producer’s share of total retail value. The evidence on widening farm gate-to-retail price spreads1 provided in this paper illustrates the plight of farmers in commodity exporting developing countries and hence provides further evidence of the urgent need for the international community to support diversiﬁcation in these economies. It should be noted that diversiﬁcation can be either horizontal diversiﬁcation to different higher value crops or vertical diversiﬁcation into higher value added activities; diversiﬁcation also does not exclude diversifying within a traditional commodity group (World Bank, 2004). ‘Resource requirements for diversiﬁcation are beyond what could possibly be mobilized at the domestic level, hence the need for establishing a ‘‘diversiﬁcation fund’’ ’ (UNCTAD, 2004b).2

5.1. Objectives The starting point of this paper is Morisset (1998) which, ﬁrst, ﬁnds an increase in the spread between world commodity prices and retail prices in consuming countries over time and, second, discusses possible explanations

1 Under the deﬁnition of farm gate-to-retail price spreads used in this paper, an increase in the farm gate-to-retail price spread over time is analytically equivalent to a decrease in the farmer’s share of retail value over time. 2 Razzaque, Grynberg and Osafa-Kwaako (2004) propose a Joint Diversiﬁcation Scheme for commodity dependent poor countries.

136

Commodity Value Chains Compression of the ﬁndings. This paper goes further using time series data of prices along entire commodity chains from raw material, in a commodity exporting developing country, to ﬁnal retail product, in a developed consuming country, in order to provide descriptive analysis of the evolution of farm gate-to-retail price spreads. Comparisons are made across several commodities—cocoa, coffee and sugar—and across countries in order to provide insight into the causes of changes in the farm gate-to-retail price spread over time. The type of explanation that is required to explain the evolution of the spreads will depend on whether trends are commodity speciﬁc and/or country speciﬁc. Sugar, in particular, has been chosen as a comparison to cocoa and coffee due to the highly distorted structure of the sugar market.

5.2. Literature Review This section reviews the areas of the commodity value chain literature that speciﬁcally concern analysis of price spreads over time and empirically test proposed explanations for the evolution of price spreads over time. Although several explanations for changes in the distribution of income along value chains over time are suggested in the literature, evidence tends to be inconclusive.

5.2.1. Background: commodity value chains In the commodity value chain literature, a commodity chain is deﬁned as a ‘network of labour and production processes whose result is a ﬁnished commodity’ (Hopkins and Wallerstein, 1986). This paper focuses on one aspect of commodity value chain analysis—the distribution of income (or value added) along the commodity chain.3 In the commodity value chain literature other concepts such as governance structures4 and institutional frameworks5 are used to explain the distribution of value along the commodity chain.

3 Gerefﬁ and Korzeniewicz (1994) identify four aspects of commodity chains (i) their inputoutput structure; (ii) their geographical coverage; (iii) their governance structure; and (iv) their institutional framework. The ﬁrst and second aspects outline the conﬁguration of the chain and therefore also outline the distribution of value added along the chain. 4 Governance refers to the sets of rules that determine the structure of the value chains, including how proﬁts and income are distributed along the value chain. In reference to value chains of manufactured goods, Gerefﬁ (1996) distinguishes ‘producer driven’ and ‘buyer driven’ value chains. Gibbon (2001) adds a third type of chain—‘international trader driven’—speciﬁcally for commodity value chains. 5 The institutional framework speciﬁes the local, national and international conditions (such as the International Commodity Agreements (ICAs) and government marketing boards, for example) that shape each activity within the chain (Ponte, 2001).

137

The Issue of Declining Commodity Prices The commodity chain can also be viewed as a series of transactions, beginning with the transfer of the raw product to the ﬁrst stage processor, and ending with the sale of the ﬁnished product to the ﬁnal consumer. These transactions may take place on a free market; they may be completely removed from the market, as in transactions between a vertically integrated TNC; or they may be structured by oligopolistic sellers and/or buyers . . . The amount of value added at each stage of the chain, and who appropriates the proﬁt from that stage, are determined by the rules governing these transactions, and by their relation to transactions at other stages. (Talbot 2002)

5.2.2. Increasing spread between world commodity prices and domestic retail prices The starting point of this paper is the empirical evidence, from the World Bank, of an increasing spread between world commodity prices and domestic retail prices in commodity importing countries (Morisset, 1998). This pattern is found across several commodities and countries.6 Although commodity exporting countries face declining prices, the expansion of ﬁnal demand in consuming countries is constrained by the lack of decline in retail prices. Associated with this increasing spread between world commodity prices and retail prices is the concentration of income in the later stages of the value chain. For example, the International Coffee Organization (ICO) estimates that in the early 1990s the value of coffee export FOB earnings by coffeeproducing countries was $10–$12 billion and the value of retail sales was $30 billion; today the value of export earnings by producing countries is $5.5 billion and the value of retail sales is $70 billion (UNCTAD, 2004b).

5.2.3. From raw material to retail product There are several studies that show the evolution of prices and the distribution of total value7 along the coffee value chain from raw material to ﬁnal retail product using data aggregated across exporting or importing countries. Talbot (1997) has data on the distribution of total value along the coffee value chain between 1974/1975 and 1994/1995 based on prices obtained by aggregating prices across all member countries of the International Coffee Organization (ICO) using a weighting based on volumes. Talbot (1997) also estimates the distribution of value added along the value chain from a speciﬁc exporting country to a speciﬁc importing country in a given year; however, 6 Morisset (1998) provides empirical evidence of the spreads between world and domestic retail price, between world and domestic wholesale prices and between domestic wholesale and retail price for the following commodity/retail product pairs: beef/beef, coffee/coffee, crude oil/fuel oil, crude oil/gasoline, rice/rice, sugar/sugar and wheat/bread, for the following importing countries: Canada, France, Germany, Italy, Japan and the United States between 1970 and 1994. 7 Total value is equal to the retail price of the ﬁnal product.

138

Commodity Value Chains Compression there is no analysis of the evolution of the distribution of income at such a disaggregated level. Recent studies on the coffee value chain, such as Ponte (2001) and Fitter and Kaplinsky (2001) update the aggregated data from Talbot (1997) to illustrate that the producer’s share of total value has been compressed over time. This paper uses time series data of prices along the value chain of a speciﬁc exporting country to a speciﬁc importing country, as in Healy (2004) which ˆ te d’Ivoire to France. In illustrates price series for the cocoa value chain from Co this paper, the time series data are used explicitly to analyse the evolution of price spreads.

5.2.4. Explanations for changes in price spreads In the literature, several explanations for the changes in the distribution of income along the value chain over time are suggested. These include: (i) production and consumption shocks (Talbot, 1997); (ii) changes in consumption habits;8 (iii) regulatory or policy changes such as the collapse of the International Commodity Agreements (ICAs);9 (iv) increasing concentration of market power in the trading, processing and retailing industries which is related to the domination of commodity chains by transnational corporations (Morisset (1998), Talbot (2002), Ponte (2001), Vorley (2003)); (v) changes in costs in production, trading, processing, marketing or retailing industries (DEFRA, 2004). However, only a few studies attempt to test proposed explanations using quantitative empirical evidence. There are several ways in which explanations are empirically examined in the literature. First, price spreads can be additively decomposed in order to identify which part of the value chain changes in price spreads occur. For example, Morisset (1998) decomposes the increasing spread between world price and domestic retail price into an increasing spread between world price and domestic wholesale price and an increasing spread between domestic wholesale price and domestic retail price. Other approaches—such as, second, examining price transmission dynamics and, third, econometrically testing the relationship between price spreads and market concentration—are more technically complex and are discussed in the following sections.

8 Fitter and Kaplinsky (2001) refer to ‘increased differentiation in consuming markets’ and Ponte (2001) refers to the ‘latte revolution’ when discussing the increase in the numbers of consumers in developed countries drinking speciality coffees : ‘it is estimated that the number of Americans drinking speciality coffees on a daily basis will grow from 20 to 27 million in 2001, up from only 7 million in 1997’ (Ponte, 2001). 9 Shepard (2004) examines price transmission in the coffee value chain and tests whether there 1989 (which is the year the International Coffee Agreement collapsed) is a structural break.

139

The Issue of Declining Commodity Prices

5.2.5. Price transmission Explanations for changes in price spreads over time often involve price transmission dynamics i.e. how ﬂuctuations in one price in the value chain are transmitted to other prices along the value chain. Morisset (1998) assumes that the direction of price transmission is from world prices to retail prices10 and based on this premise provides empirical evidence that upward movements in world price are better transmitted to retail prices than downward movements in world price. This ‘asymmetric transmission’ argument, which occurs often in the commodity value chain literature, is used to explain the increasing spread between world commodity prices and domestic retail prices. There is, though, no a priori reason for the assumption regarding the direction of price transmission. Price transmission may be in either or even both directions along the value chain. There may even be different transmission dynamics along different sections of the value chain from raw material to retail product.11 Econometric tests, such as the Granger causality test,12 are necessary to determine directions of causality between prices along the value chain.13 Testing for asymmetric transmission involves testing whether there is a statistically signiﬁcant difference in the extent to which upwards and downwards movements in a price—namely the world price in Morisset (1998)—are transmitted to prices further along the value chain—namely the domestic retail price in Morisset (1998). Morisset’s methodology for examining price transmission is based on Mundlak and Larson (1992) which estimates price elasticities assuming a direct relationship between world market prices and domestic prices. Other price transmission studies use an alternative methodology involving error correction speciﬁcations following Engle and Granger (1987). In comparison to Mundlak and Larson’s static model, error correction models are dynamic allowing domestic prices to adjust to their long-term equilibrium in the period following a change in the world price. DEFRA (2004) uses an error correction model to test for asymmetric transmission in agricultural value chains within EU countries (i.e. the producer and retail price are both for the same country) and ﬁnds no systematic evidence of the existence of asymmetric transmission. 10 In technical terms, it is assumed that there is a long run relationship between world price and retail price after controlling for exogenous variables such as production and consumption shocks. 11 Price transmission is likely to be related to ‘governance structures’ that have been discussed in footnote 8. Talbot (2002) observes that there can be different governance structures along different segments of value chains. 12 This tests whether transmission is (i) bottom-up (i.e. from producer prices to retail price), (ii) top-down (i.e. from retail prices to producer prices), (iii) in both directions or (iv) very weak. 13 See DEFRA (2004) Annex 8 for an investigation into the direction of causality between retail prices and farm gate prices in EU agricultural value chains and Shepard (2004) for an investigation into price transmission through the coffee value chain. Both studies ﬁnd that transmission occurs both up and down the value chain.

140

Commodity Value Chains Compression

5.2.6. Market power explanations Morisset (1998) suggests a market power explanation for the asymmetric transmission of downwards and upwards movements in world commodity prices to domestic price and hence the increasing spread between world commodity prices and domestic retail prices. Likewise, in much of the value chain literature it is assumed that the increase in share of total value in later stages of the value chain is due to concentration of market power in the later stages of the value chain. There is much anecdotal evidence which indicates that this is the case; for example, See Ponte (2001), Fitter and Kaplinsky (2001), and Vorley (2003) for discussion of market concentration at different stages of activity—such as production, trading, processing, retailing—along value chains. However, the link between market concentration and the increase in concentration in the later stages of the value chain has not been proven rigorously in the literature. Morisset (1998) either rejects alternative explanations or argues them to be insufﬁcient to explain the degree of increase in the spread between world commodity prices and ﬁnal retail prices. A ‘trade restrictions’ explanation— that changes in the spread are the result of trade restrictions in consumer markets—is rejected as the evolution of the spread between world and retail prices is similar across consuming countries despite differences in trade policies. Variants of ‘bottleneck’ explanations—that increasing costs of marketing, processing, distribution, or retail are the cause of the increasing spread— are argued to be insufﬁcient explanations for the magnitude of the increase in the spread between world and retail prices. Morisset (1998) argues that the weight of processing costs, in total retail cost, would have to be disproportionately large in comparison to the weight of the world commodity price in order to explain the observed increase in average spread between world commodity price and retail price. There are several issues that arise relating to testing whether there is a link between market power and price spreads. First, a suitable deﬁnition of market power must be used and, second, other variables that could impact price spreads must be taken into account. DEFRA (2004) tests a hypothesized relationship between concentration of market power in retailing and farm gate-toretail price spreads from EU agricultural value chains for a number of broad food groups.14 First, a simple correlation analysis ﬁnds (i) only weak correlation between farm gate-to-retail price spreads and concentration in retailing and (ii) a generally negative relationship between changes in farm gate-to-retail price spreads and changes in concentration in retailing. However, correlation analysis does not take other variables into account that could also be impacting 14 The broad food groups are: (i) wheat; (ii) red meat; (iii) poultry; (iv) fruits and vegetables. Data are used from Austria, Denmark, France, Germany, Ireland, Italy, the Netherlands, Spain and the UK.

141

The Issue of Declining Commodity Prices price spreads. Second, an econometric model which takes other variables, such as (i) costs along the supply chain, (ii) demand and supply conditions, (iii) European Union intervention prices under the Common Agricultural Policy and (iv) the exchange rate where relevant, into account is used to test the hypothesis; this method fails to establish a statistically signiﬁcant relationship between concentration in retailing (used as a proxy for market power) and farm gate-to-retail price spreads, with the exception of fruit and vegetables.

5.2.7. Conclusion There is no consensus, in the commodity value chain literature, on what causes the evolution of price spreads over time: determining what causes the evolution of price spreads is likely to be a complex and arduous task. In this paper, cross-country and cross-commodity comparisons are made in order to provide both a description of evolution of farm gate-to-retail price spreads and to provide insight into the type of explanation that is needed to explain the changes in these spreads over time.

5.3. Data This chapter focuses on value chains from commodity exporting developing countries to commodity importing developed countries.15 Data for the following prices is collected: (i) the price paid to the farmer at the farm gate for the raw material in a commodity exporting developing country; (ii) the free on board (fob)16 export price of the commodity exporting developing country;17 and (iii) the retail price of a ﬁnal product in a developed consuming country. Clearly, any results rely on both the suitability and consistency of the time series data used. When examining the value chains of commodities, it is not always clear which form of the commodity should be analysed at each stage; this raises several issues relating to comparability of price spreads across both countries and commodities. There are normally several distinct processed forms and/or end uses of a given raw material. Often, the commodity, such as cocoa, will only be part of the content of the product of end use, such as chocolate or chocolate confectionery. This will have implications for the comparability of value chains across commodities. Where possible, a ﬁnal retail product, such 15 For sugar, value chains from commodity exporting developed countries are also considered in order to provide comparisons to value chains from developing countries which export to the EU under preferential trading agreements i.e. the Sugar Protocol. 16 The free on board price is the price on goods delivered onto the ship at the port of export. 17 Unlike Morisset (1998), this paper uses export fob prices, derived from export fob volumes and values, rather than using world market prices (or a composite indicator price) as a proxy for the prices paid to commodity exporting developing countries.

142

Commodity Value Chains Compression as granulated sugar, that consists primarily of the commodity, is considered. The main types of end use for a commodity may also differ across producing countries; for example, much of sugar cane produced in Brazil is used for domestic ethanol production rather than being exported in the form of raw sugar. This has implications for the comparability of value chains for a given commodity across countries. A commodity may be exported in several different forms or stages of processing; countries often export both raw and white sugar, for example. The form of each commodity that has been chosen to be examined at each stage of the value chain considered in this paper—namely farm gate, export and retail—is outlined and discussed below. The sources and descriptions of the prices used in this paper are listed in Appendix 5.1 and discussed below. As the price spreads examined in this paper involve prices from both an exporting and an importing country, the accuracy of the data will depend on the accuracy of the exchange rates from the IMF International Financial Statistics; it is assumed that the exchange rate is neither undervalued nor overvalued.

5.3.1. Coffee The coffee price time series are obtained from the International Coffee Organization (ICO) which collects data from its member countries. The ICO has monthly data from 1984 to 2003; in this paper yearly averages are used to take seasonal variation into account. The farm gate price is the price of coffee beans, in the form most often purchased from the grower, paid to the grower.18 There are two main types of coffee for which the ICO has farm gate price time series: arabica and robusta. Although many countries export both types of coffee, in this paper the farm gate price of the major type of coffee exported is used for each exporting country—arabica for Ethiopia, Kenya, Papua New Guinea and Tanzania and robusta for Cameroon. A weighting could have been used to provide a proxy of the average farm gate price; however, this weighting would be arbitrary as the ratio of arabica to robusta coffee exports varies across both countries and time. As robusta coffee is generally cheaper to produce than arabica coffee, the farm gate-to-retail price spread19 will be greater in absolute magnitude if the farm gate price of robusta rather than arabica is used. Therefore, using a farm gate price series of robusta coffee provides an upper bound to the farm gateto-retail price spread and using a farm gate price series of arabica coffee provides a lower bound to the farm gate-to-retail price spread. 18 See ‘Data Concepts and Variable Used in the Statistics of the Organization’, ICO (2003), for details about ICO variables. Processing the cherries to parchment coffee is usually done by the growers who tend to be small to medium sized farmers in most producing countries. Processing parchment coffee into green coffee is usually done by larger growers or other actors as it requires more sophisticated machinery (Talbot, 2002). 19 Using the deﬁnition of farm gate-to-retail price spread given below.

143

The Issue of Declining Commodity Prices The export fob price is the export unit value of all green bean exports (i.e. both arabica and robusta exports). This is derived by the ICO by dividing the total value of coffee exports by total volume of exports. The UK retail price is based on the retail price of 100 g of instant coffee. The ICO has retail price data based on the retail price of 1 kg of roasted coffee in importing countries other than the UK; as the trends of the price spreads are similar, whether the retail price series of roasted coffee or instant coffee used the value chains with the UK as the importing country will be discussed in this paper.

5.3.2. Cocoa This paper uses farm gate and export fob prices for cocoa beans. Most of cocoa trade on the world market is in the form of cocoa beans although cocoa butter and powder are also traded (Talbot, 2002). These yearly farm gate and export fob prices from 1983 to 2003 for Ghana have been obtained from the International Cocoa Organization (ICCO). A UK price index of chocolate products is used as the retail price in the cocoa value chain. There are several processed cocoa products such as cocoa liquor, cocoa paste, cocoa butter, cocoa powder and cocoa cake. Apart from cocoa powder, these processed products are not ﬁnal retail products but are used to make other products: cocoa powder is used in chocolate ﬂavoured desserts; cocoa butter is used to make chocolate and also in cosmetics such as moisturizing cream and soap; cocoa liquor is used to make chocolate. Chocolate, the most obvious ﬁnal retail cocoa product, is made from both cocoa liquor and butter and other ingredients such as sugar; the ICCO estimates that about two thirds of cocoa production is used to make chocolate. Chocolate is used to make chocolate confectionery as well as being a ﬁnal retail product. The chocolate retail price series used has been compiled by the ICCO using data from the Biscuit, Cake, Chocolate and Confectionery Association (BCCCA)—it is the average retail price of chocolate products.

5.3.3. Sugar The sugar farm gate and export fob price series for Australia, Brazil, Fiji, Mauritius and Thailand, from 1984 to 2003, are obtained from LMC International, a commodities consultancy that is the premier source of sugar data. Retail price series for the UK and US are obtained from National Statistics Ofﬁces. The farm gate price is the price of sugarcane paid to the growers. The majority of the world’s sugar—about 83 per cent—is produced from sugar cane with the remainder being produced from sugar beet (Vorley, 2003). Sugarcane is processed into white sugar in two steps: ﬁrst, sugar cane is processed into raw sugar and then the raw sugar is reﬁned into white sugar; sugar beet is processed straight into white sugar.

144

Commodity Value Chains Compression The export fob price is the export price of raw sugar. As sugar is exported to the EU and US under preferential trading agreements, which will be discussed in section 5.5, the export fob price varies according to the import destination. It should be remembered that most sugar exporters export both raw sugar and white sugar: if export prices impact on other prices, such as the sugar cane price, then it is likely that the prices of both raw and white sugar would be signiﬁcant. The retail price is for white granulated sugar. Other sugar products include brown sugar, syrup and products made from by-products such as molasses20 and bagasse.21 For some countries, by-products are major end uses of sugar; a signiﬁcant share of Brazilian sugar, for example, is used in the production of ethanol. If the price of the product of end use is signiﬁcant then it is likely that the price of ethanol would also be signiﬁcant.

5.3.4. Deﬁnition of price spread In this chapter price spreads are calculated in order explicitly to analyse trends. Price spreads are deﬁned as follows: the farm gate-to-retail price spread, for example, is the retail price divided by the farm gate price. In other words, the farm gate-to-retail price spread is the inverse of the share of total value received by farmers. An increasing trend in the farm gate-to-retail price spread (which is retail price/farm gate price) over time indicates that there has been compression of the farmer’s share of retail value over this time period. Price spreads are deﬁned in this way for two reasons. First, a relative rather than an absolute deﬁnition allows cross country and cross commodity comparisons to be made. Second, it is convenient to analyse the decomposition of the farm gate-to-retail price spread into (i) the farm gate-to-export fob price spread and (ii) the export fob-to-retail price spread as the farm gate-to-export fob and export fob-to-retail price spreads multiply to give the farm gate-toretail price spread or, alternatively, the log of the farm gate-to-export fob price spread and the log of the export fob-to-retail price spread adds to give the log of the farm gate-to-retail price spread.

5.4. Coffee and Cocoa 5.4.1. Coffee: farm gate-to-retail price spread As discussed earlier, existing studies indicate that at an aggregate level there has been compression of the farmer’s share of total value in the coffee value chain. In this paper, a twenty year time series is used to calculate farm gate-toretail price spreads for speciﬁc exporting and importing countries. 20 21

Molasses can be used as cattle feed or distilled to produce ethanol. Bagasse is used as fuel.

145

The Issue of Declining Commodity Prices From Figures 5.1 to 5.5, which show the logs of farm gate-to-retail price spreads, it can be seen that there is an increasing trend in the farm gate-toretail price spread22 for the following coffee value chains: Cameroon-UK, Ethiopia-UK, Kenya-UK, Papua New Guinea-UK, and Tanzania-UK. For example, for the Cameroon-UK coffee value chain, the trend of the log of the farm gate-to-retail price spread increases by 0.03 per year between 1984 and 2003.

5.4.2. Coffee: decomposition of farm gate-to-retail price spread As discussed in section 3, the log of the farm gate-to-retail price spread can be additively decomposed into the log of the farm gate-to-export fob price spread and the log of the export fob-to-retail price spread. For the commodity exporting countries considered, the trends in price spreads are similar whether the UK or the US is the importing country (and instant or roasted coffee are the respective retail products). In the discussion that follows, the value chains for countries exporting to the UK will be examined. The log of the export fob-to-retail price spread has an upward trend for Cameroon-UK, Ethiopia-UK, Kenya-UK, Papua New Guinea-UK, and Tanzania-UK; this is analytically equivalent to the increase in spread between world commodity prices and retail prices documented in Morisset (1998). The log of the farm gate-to-export fob price spread has a downward trend for Ethiopia and a ﬂat trend for Cameroon, Kenya and Papua New Guinea, and Tanzania. Despite an absence of compression of the farmer’s share of export price from 1984 to 2003, there has been compression of the farmer’s share of retail price over the same time period. In order to ascertain the cause of the upward trend in farm gate-to-retail prices, the upward trend in export fob-to-retail price spreads must be explained. For Ethiopia-UK, the farmer’s share of total value is being compressed despite the fact that the farmer’s share of export price is increasing. This result mirrors the observation that, in terms of the farmer’s share of total value, farmers are not beneﬁting from a decrease in the margins between farm gate price and export fob price in commodity exporting developing countries over time (Fitter and Kaplinsky 2001). Finding a similar pattern of trends across value chains in different exporting and importing countries suggests that the explanation for the upward trend in farm gate-to-retail price spreads is not country speciﬁc.

22

As the log function is a monotonic transformation, an upward trend in the log of a price spread is equivalent to an upwards trend in the price spread.

146

Commodity Value Chains Compression

5.4.3. Cocoa The log price spreads from the Ghana-UK cocoa value chain, shown in Figure 5.6, have the following features: the log farm gate-to-retail price spread has an upward trend which is decomposed into a log farm gate-to-export fob price spread with a downward trend and a log export fob-to-retail price spread with an upward trend. The farmer’s share of total value is being compressed over time despite an increase in the farmer’s share of export price over the same time period.

5.4.4. Cross commodity comparisons For both the coffee and cocoa value chains the farm gate-to-retail price spreads have an increasing trend. This long run increase in the farm gate-to-retail price spread indicates that the farmer’s share of total value is being compressed over time. This compression is occurring despite the absence of compression of the farmer’s share of export price over the same time period for both commodities, suggesting that explanations for an increasing farm gate-to-retail price spread are not commodity speciﬁc.

5.5. Sugar 5.5.1. Background The world sugar market is highly distorted due to sugar policies in both exporting and importing countries. The importing countries considered in this paper, the UK and US, are also sugar exporters and have complex and controversial sugar policies to protect their domestic farmers. Countries exporting to the UK and US beneﬁt to differing degrees from preferential access to these markets. As there are also government interventions in the producing and exporting countries considered, such as Brazil and Australia, caution must be used when making comparisons across value chains in different countries. Under the European Union’s Common Market Organization (CMO) there is price support for set quotas of A and B sugar which is either sold internally at the EU guaranteed intervention price or exported with refunds based on the difference between the EU price and the world price. Any sugar produced domestically in excess of these quotas is known as C sugar and must be exported without subsidy. Under preferential trading agreements there are quotas for mainly ACP countries to export raw sugar to the EU at the guaranteed price; there are high levels of protection, in the form of import tariffs, to deter non-preferential imports. Preferential sugar imports from ACP countries have taken place under the Sugar Protocol, since the 1975 Lome´ Convention

147

The Issue of Declining Commodity Prices Table 5.1. Breakdown of 2003 Raw Sugar Sales Percentage of total raw sugar sales in 2003 Country Australia Brazil Fiji Mauritius Thailand

Domestic sales Preferential exports to EU Preferential exports to US Exports to rest of world 55 0 0 0 0

0 0.2 77 98 0

1 2 3 2 0.05

44 98 19 0 99.5

between the EU and ACP countries, and under Special Preferential Sugar (SPS) Agreements since 1995; in addition, the EU has preferential sugar imports from India and, since 1995, imports of Most Favoured Nation sugar from other countries, such as Brazil. Under US sugar policy: loans provide a minimum price for domestically produced sugar; raw sugar and white sugar, for ﬁnal consumption within the US, is imported under the Tariff Rate Quota (TRQ) system which effectively limits imports in order to guarantee a high domestic price. The TRQ is a two tiered tariff: a lower tariff is charged on imports within quotas; a prohibitively high tariff is charged on imports in excess of these quotas. The quotas, which are allocated to over 40 countries based on exports to the US between 1975 and 1985, give countries preferential sugar trading agreements with the US. Under the sugar re-export programme, sugar which is used in the production of products which are exported can be imported at world price. In this paper, as value chains are examined from raw material in an exporting country to ﬁnal retail product in an importing country, sugar imported under the Sugar Reexport Programme at world price is not taken into consideration. The export fob price for countries exporting to the EU and US is higher than the export fob price for non-preferential exports to the world market. The degree to which a country exports under preferential trading agreements varies across exporting countries. Table 5.1 shows a snapshot of the 2003 breakdown of total raw sugar sales—for Australia, Brazil, Fiji, Mauritius and Thailand—into domestic sales, exports to the EU and the US under preferential trading agreements and exports to the rest of the world. Price time series for two ACP countries—Fiji and Mauritius—which export raw sugar to the EU under the Sugar Protocol and SPS are used in this paper. In order to make suitable comparisons between countries that export a large percentage of raw sugar under preferential trading agreements and countries that export only a small percentage of total sugar under preferential trading agreements, FijiUS and Mauritius-US price spreads will be compared to Australia-US, Brazil-US and Thailand-US price spreads. Although all these countries only export a couple of per cent of their total raw sugar sales to the US, this comparison could yield

148

Commodity Value Chains Compression some insight into whether the degree to which a country exports raw sugar under preferential trading agreements affects compression.

5.5.2. Absence of compression It is striking that there is a lack of compression of the farmer’s share of total value for Fiji-EU and Fiji-US; the trend of the log farm gate-to-retail price spreads in Figures 5.11 and 5.12 is decreasing indicating that Fijian farmer has been gaining an increasing share of retail value over time. This seems to be a marked difference from the increasing trend for the log farm gate-to-retail price spreads from both the cocoa and coffee value chains.

5.5.3. Comparing the farm gate-to-retail price spread between protocol and non-protocol countries Figure 5.7 shows the evolution of the farm gate-to-retail price spreads for Australia-US, Brazil-US, Fiji-US, Mauritius-US and Thailand-US. There is a clear difference between the trends for Fiji and Mauritius—which export over three-quarters of their total volume of raw sugar sales under preferential trading agreements—and the trends for Australia, Brazil and Thailand—who export less than three per cent of their total volume of raw sugar sales under preferential trading agreements. The extent of compression of farmer’s share of total value is much greater for Brazil-US, Australia-US and Thailand-US than Mauritius-US. As discussed above, for Fiji-US, there is no compression; the farm gate-to-retail price spread has a declining trend. This suggests that there may be transmission from average export fob price to farm gate price. The higher the percentage of the total volume of raw sugar sales of a country that is exported under preferential trading agreements (i.e. exported at a guaranteed price that is higher than the world sugar price), the higher the average export fob price of raw sugar for that country. For countries that export the majority of their total raw sugar sales on the world market, the average export fob price will be around the level of the price of sugar on the world market. The evidence seems to suggest that preferential trading agreements insulate countries such as Fiji and to some extent Mauritius from pressures that lead to compression of the farmer’s share of total value.

5.5.4. Decomposition of the farm gate-to-retail price spread Figures 5.9 to 5.15 show the decomposition of the log farm gate-to-retail price spreads for the sugar value chain. The log export fob-to-retail price spread and log farm gate-to-export fob price spreads for the Mauritius-EU, Mauritius-US, Fiji-EU, and Fiji-US value chains have the same trends as those from the coffee and cocoa

149

The Issue of Declining Commodity Prices value chains: the log farm gate-to-export fob price spreads have downward or ﬂat trends; the log export fob-to-retail price spreads have upward trends. In comparison to the log farm gate-to-export fob price spreads from the FijiUS and Mauritius-US value chains, the log farm gate-to-export fob price spreads from the Australia-US, Brazil-US, and Thailand-US value chains have upward trends. For Australia, Brazil and Thailand the margin between farm gate and export price is an increasing share of export value. For the sugar value chains, the log of the farm gate-to-export fob price spread makes a larger contribution to the log of the farm gate-to-retail price spread than the log of the export fob-to-retail price spread. This is the reverse of the pattern for the coffee value chains.

5.6. Conclusion There is a clear increasing trend in the farm gate-to-retail price spread (i.e. a decreasing trend in the farmer’s share of total value) over time for the coffee and cocoa value chains. This is also the case for the farm gate-to-retail price spreads from sugar value chains originating in countries that export most of their sugar on the world market. This provides evidence that the farmer’s share of total value is being compressed across several commodities. For the coffee and cocoa value chains, there is a pattern in the decomposition of the farm gate-to-retail price spread both across countries and across the commodities: the increase in farm gate-to-retail price spread over time appears to be caused be an increase in export fob-to-retail price spreads. This suggests that the explanation for the observed evolution in these price spreads is neither country nor commodity speciﬁc. It is particularly interesting that the evidence suggests that the farmer’s share of value added in sugar exporting countries which are insulated to a large extent from the world sugar market by preferential trading agreements with the EU is not being compressed to the same extent as farmers in countries which do not beneﬁt from the Sugar Protocol. This ﬁnding has serious implications for farmers facing preference erosion: without preferential treatment they will no longer be insulated from pressures which seem to be leading to the compression of the farmer’s share of total value in commodity value chains.

150

Commodity Value Chains Compression Coffee: Cameroon-UK 2 1.8

log price spread

1.6 log (retail price/farm gate price)

1.4 1.2

log (retail price/ export fob price)

1 0.8

log (export fob price/ farm gate price)

0.6 0.4 0.2 00 20 02

98

20

96

19

94

19

92

19

90

19

19

86

88

19

19

19

84

0

Year Figure 5.1. Coffee: Cameroon-UK

Coffee: Ethiopia-UK 1.8 1.6

log price spread

1.4 log (retail price/farm gate price)

1.2 1

log (retail price/ export fob price)

0.8

log (export fob price/ farm gate price)

0.6 0.4 0.2

86 19 88 19 90 19 92 19 94 19 96 19 98 20 00 20 02

19

19

84

0

Year Figure 5.2. Coffee: Ethiopia-UK

151

The Issue of Declining Commodity Prices

Coffee: Kenya-UK 1.8 1.6 1.4 log price spread

1.2

log (retail price/farm gate price)

1

log (retail price/ export fob price)

0.8 0.6

log (export fob price/ farm gate price)

0.4 0.2

00 20 02

98

20

19

94

96

19

92

19

90

19

19

88

86

19

19

19

84

0

Year Figure 5.3. Coffee: Kenya-UK

Coffee: PNG-UK 1.8 1.6 1.4 log price spread

1.2

log (retail price/farm gate price)

1

log (retail price/ export fob price)

0.8 0.6

log (export fob price/ farm gate price)

0.4 0.2

90 19 92 19 94 19 96 19 98 20 00 20 02

19

88

19

19

19

84

−0.2

86

0

Year Figure 5.4. Coffee: PNG-UK

152

Commodity Value Chains Compression

Coffee: Tanzania-UK 2 1.8

log price spread

1.6 1.4

log (retail price/farm gate price)

1.2

log (retail price/ export fob price)

1 0.8

log (export fob price/ farm gate price)

0.6 0.4 0.2 86 19 88 19 90 19 92 19 94 19 96 19 98 20 00 20 02

19

19

84

0

Year Figure 5.5. Coffee: Tanzania-UK

Cocoa: Ghana-UK 1.6

log price spread

1.4 1.2

log (retail price/farm gate price) log (retail price/export fob price) log (export fob price/ farm gate price)

1 0.8 0.6 0.4 0.2

19

83 19 /84 85 19 /86 87 19 /88 89 19 /90 91 19 /92 93 19 /94 95 19 /96 97 19 /98 99 20 /00 01 /0 2

0

Year Figure 5.6. Coffee: Ghana-UK

153

The Issue of Declining Commodity Prices Sugar: farm gate-to-retail price spreads 90 80

Price spread

70

US retail/Mauritius farm gate

60

US retail/Fiji farm gate

50 40 30

US retail/Thailand farm gate

20

US retail/Brazil farm gate

10

US retail/Australia farm gate

19

84 19 86 19 88 19 90 19 92 19 94 19 96 19 98 20 00 20 02 20 04

0

Year Figure 5.7. Sugar: farm gate-to-retail price spreads

Sugar: farm gate-to-retail price spreads 90 80

Price spread

70 UK retail/Mauritius farm gate

60 50 40

UK retail/Fiji farm gate

30

UK retail/Brazil farm gate

20 10 19 84 19 86 19 88 19 90 19 92 19 94 19 96 19 98 20 00 20 02 20 04

0

Year Figure 5.8. Sugar: farm gate-to-retail price spreads

154

Commodity Value Chains Compression

log (retail price/ farm gate price) log (retail price/ export fob price)

04

02

20

00

20

98

20

96

19

94

19

92

19

90

19

88

19

19

19

19

86

log (export fob price/ farm gate price)

84

Log price spread

Sugar: Mauritius-US 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

Year Figure 5.9. Sugar: Mauritius-US

19

19

19

88 19 90 19 92 19 94 19 96 19 98 20 00 20 02 20 04

log (retail price/farm gate price) log (retail price/export fob price) log (export fob price/ farm gate price)

84 86

Log price spread

Sugar: Mauritius-EU 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

Year Figure 5.10. Sugar: Mauritius-EU

155

The Issue of Declining Commodity Prices

04

02

20

00

20

98

20

96

19

94

19

92

19

90

19

88

19

19

19

19

86

log (retail price/farm gate price) log (retail price/export fob price) log (export fob price/ farm gate price)

84

log price spread

Sugar: Fiji-US 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

Year Figure 5.11. Sugar: Fiji-US

log (retail price/farm gate price) log (retail price/export fob price) log (export fob price/ farm gate price)

19 84 19 86 19 88 19 90 19 92 19 94 19 96 19 98 20 00 20 02 20 04

log price spread

Sugar: Fiji-EU 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

Year Figure 5.12. Sugar: Fiji-EU

156

Commodity Value Chains Compression

04

02

20

00

20

98

20

96

19

94

19

92

19

90

19

88

19

19

19

19

86

log (retail price/ farm gate price) log (retail price/export fob price) log (export fob price/ farm gate price)

84

log price spread

Sugar: Brazil-US 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

Year Figure 5.13. Sugar: Brazil-US

log (retail price/ farm gate price) log (retail price/ export fob price) log (export fob price/ farm gate price)

19

84 19 86 19 88 19 90 19 92 19 94 19 96 19 98 20 00 20 02 20 04

log price spread

Sugar: Thailand-US 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

Year Figure 5.14. Sugar: Thailand-US

157

The Issue of Declining Commodity Prices

01

99

97

03 20

20

19

19

95 19

93 19

19

19

91

log (retail price/farm gate price) log (retail price/ export fob price) log (export fob price/ farm gate price)

89

log price spread

Sugar: Australia-US 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

Year Figure 5.15. Sugar: Australia-US

Appendix 5.1. Commodity

Price

Country

Time period

Details

Data Source

Sugar

Farm gate price

Australia Brazil Fiji Mauritius Thailand Australia

1989–2003 1984–2003 1984–2003 1984–2004 1984–2003 1984–2004

Farm gate price of sugar cane

LMC International

LMC International

Brazil

1984–2004

Fiji

1984–2004

Mauritius

1984–2004

Thailand

1984–2004

US

1984–2004

Export price under US TRQ; Export price under US TRQ; Export price under MFN to EU post 1995 Export price under US TRQ; Export price under Sugar Protocol and SPS to EU Export price under US TRQ; Export price under Sugar Protocol and SPS to EU Export price under US TRQ; Based on retail price of granulated sugar

Export FOB price for preferential sales (EU and USA)

Retail price

US Department of Agriculture (USDA)

(Continued )

158

Commodity Value Chains Compression Appendix 5.1. (Continued ) Commodity

Coffee

Price

Country

Time period

Details

Data Source

UK

1987–2004

Ofﬁce of National Statistics (ONS)

Cameroon

1984–2003

Ethiopia

1984–2003

Kenya

1984–2003

Papua New Guinea

1984–2003

Tanzania

1984–2003

Cameroon

1984–2003

Based on retail price of 1 kg of granulated sugar Farm gate price of robusta coffee Farm gate price of arabica coffee Farm gate price of arabica coffee Farm gate price of arabica coffee Farm gate price of arabica coffee Unit value of coffee exports (both arabica and robusta). Derived by dividing total value (FOB) of coffee exports by total volume of coffee exports in Green Bean Equivalent.

Ethiopia Kenya Papua New Guinea Tanzania US

1984–2003 1984–2003 1984–2003 1984–2003 1984–2003

UK

1984–2003

Farm gate price

Ghana

1984–2003

Export FOB price Retail price

Ghana

1984–2003

UK

1980–1998

Farm gate price

Export FOB price

Retail price

Cocoa

Based on retail price of 1 kg roasted coffee Based on retail price of 100 g soluble coffee Farm gate price of cocoa beans Export price of cocoa beans Average retail price of chocolate products compiled by ICCO using data from BCCCA. This time series has been extrapolated ﬁve years (the R squared of a best ﬁt linear trend is 0.99).

International Coffee Organization (ICO)

International Cocoa Organization (ICCO) ICCO, Biscuit, Cake, Confectionery and Chocolate Association (BCCCA), ONS

This page intentionally left blank

PART II The Implications of Declining Commodity Prices

This page intentionally left blank

6 Estimating Foreign Exchange Loss due to Declining Commodity Prices Mohammad A. Razzaque, Philip Osafa-Kwaako, and Roman Grynberg

The sustained downward trend in relative prices of primary commodities has resulted in signiﬁcant foreign exchange losses to most of the developing countries that rely on them for exports. Since export earnings determine a country’s capacity to import, secular deterioration in the net barter terms of trade will imply reduced purchasing power of a given volume of commodity.1 In other words, there will be a net loss of foreign exchange from commodity exports. This chapter estimates the magnitude of these losses incurred by LDCs, HIPCs, and small states.

6.1. Method The concept of foreign exchange loss should be considered in terms of the difference between the volume of primary exports and its purchasing power taking into consideration changes in commodity prices relative to other products.2 Given a ﬁxed quantity of commodity, the amount of forgone earnings can be estimated if the trend decline rate in prices of commodities relative to manufactured goods is known a priori.3 Practical implementation of this 1 Many commodity-dependent developing countries, however, receive foreign aid, which contributes to their capacity to import. Services exports, including remittances from nationals working abroad, are also a source of import ﬁnance. Nevertheless, earnings from exports of merchandise goods are the single most important factor in determining import capacity for the vast majority of commodity-dependent developing countries. 2 Manufactured goods and fuels are principal import items in LDCs, SVs and HIPCs. Hence, the purchasing power of primary exports from these countries is mostly inﬂuenced by the volume of commodity exports and the import prices of non-primary goods. 3 Suppose the quantity of exports is ﬁxed at 50 units and per unit real price in 1980 is US$100—so that total export revenue is US$5000. A trend decline rate of 1 per cent per annum

163

The Implications of Declining Commodity Prices method will result in a number of problems. First, the quantity is likely to vary considerably, and since commodity-dependent countries are subject to various exogenous shocks, such as natural calamities, ﬁxing quantity at some representative level may not be easy. Alternatively, however, one can allow for variation in quantity and the loss can be computed by comparing the actual revenues in each period with hypothetical earnings based on the real price in the reference period. Another important problem is that the estimation of the trend rate is sensitive to the choice of sample period and, consequently, the representative rate of decline may depend on individual researchers.4 Most importantly, the export baskets of LDCs, HIPCs, and SVs comprise several commodities, which makes the task of estimating commodity-speciﬁc losses for all countries virtually impossible.5 Due to these problems, in this chapter we follow the methodology used by Maizels (1992) and UNCTAD (2002a) in calculating the foreign exchange loss from commodity exports. The methodology is outlined below. Suppose a country exports only primary commodities. The value of its real exports at some base year prices in each year, t, may be stated as: n X Vxit Pxi i¼1

(1)

where Vxi refers to the value of exports of commodity i (¼ 1, 2,. . . . , n) in current prices, and Pxi is the corresponding export unit value index with the pre-speciﬁed base year. The purchasing power of exports in any given year is expressed as: n X Vxit Pmi i¼1

(2)

where Pmi is the corresponding import unit value index. The foreign exchange loss can then be computed simply as the difference between equations (1) and (2).6 would cause the real price to fall to US$81.79 in 2000, resulting in a foreign exchange loss of about US$910 for that year. 4 In Chapter 3, it has been illustrated that the use of a very long-time series, e.g. 1900–2001, results in much lower trend decline rates than a relatively short sample. 5 Also, apart from the tedious computational burden, for many commodities price information is not available. Similarly, the information on quantity of all primary commodities for all countries is not available. 6 An alternative interpretation of equations (1) and (2) is obtained by decomposing each expression into a more reduced form, and expressing it in real terms—i.e. in terms of real imports of individual countries. Consider the hypothetical case in which a country exports one primary good, x, and imports are comprised solely of manufactures, m. In this case, Equation (1) may then be restated as:

164

Estimating Foreign Exchange Loss There are a number of issues relating to equations (1) and (2). First of all, it is unrealistic to assume that all countries in our sample rely solely on commodities for exports. To overcome this problem we consider the value of exports emanating from primary commodities only. This reﬁned deﬁnition not only excludes all earnings from manufactured goods but also leaves out receipts from fuel exports. The UNCTAD database provides such country-speciﬁc information on primary exports. The choice of unit value indices is another factor in using the methodology. For Px in equation (1), one should ideally construct commodity price indices that reﬂect the composition of individual countries’ primary exports. This requires information on the unit value of all commodity export items. Unfortunately, such detailed country-speciﬁc information is not available.7 Alternatively, we use country-speciﬁc export unit value indices, as available in UNCTAD (2002b) as proxies for Px in equation (1).8 For Pm in equation (2), we have a more appropriate index, which is the country-speciﬁc import unit value index as available in UNCTAD (2002b). The most critical problem is associated with the choice of base year. Since relative commodity prices have experienced secular deterioration, a choice of base year in the relatively early period in the sample will show a much higher loss than a later period in the sample. Maizels (1992) chooses 1980 as the base year, which is often criticized on the grounds that it was a year associated with high commodity prices. In UNCTAD (2002a), therefore, 1986 is used as the

Vxt Pxindex

¼

Ptx Qtx Ptx Pb

¼ Pbm

Pb Qtx Pbm

(3)

where P and Q represent prices and quantities, respectively. Thus, in any year t, equation (1) represents the purchasing power of exports, Qtx evaluated at base year prices for commodities Pb , and for manufactures Pbm . Equation (2) may also be restated as: x x Vxit Px Q x P Q ¼ t t ¼ Pbm t t Pmin dex Ptm =Pbm Ptm

(4)

which represents the purchasing power of exports, Qtx , evaluated at current prices for commodities Ptx , and manufactures Ptm . The difference between (2) and (1) therefore represents the difference in import capacity of the two countries evaluated at base year prices of manufactures, Pbm . Using the same method, it is possible to calculate the magnitude of commodityspeciﬁc foreign exchange losses. The unit value of exports in this case is the price of the relevant primary commodity. 7 Deaton and Miller (1996) construct a geometrically weighted export price index for a number of countries. However, the procedure cannot be applied to all countries in our sample due to data unavailability. Ideally, an import unit value index should exclude all primary commodities, but there is no study that attempts to construct such a unit value index. 8 The use of an overall export unit value index is not unusual. Maizels (1992) and UNCTAD (2002b) also applied the same index in their estimation. Nevertheless, it should be pointed out that in the event of changes in export composition, e.g. diversiﬁcation away from primary commodities, the use of the aggregate unit value of exports might not adequately reﬂect the real value of primary exports in equation (1).

165

The Implications of Declining Commodity Prices base year. Figure 6.1 shows that the relative aggregate (composite) commodity price index, measured on the left vertical axis, is much the same as the average for 1976–79 and is not signiﬁcantly different from the average for the 1960s. On the other hand, the year 1986 clearly represents a period with very low relative prices and, therefore, will certainly undermine the magnitude of foreign exchange loss for commodity-exporting poor countries. To provide a more balanced picture, we have chosen the average for the years 1984, 1985, and 1986 as the reference base prices. Some indication will also be given as regards the effects of using 1980 as the base year. Finally, it is worth pointing out that the formulation for estimation of losses, as outlined above, does not explicitly reﬂect the effect of a secular declining trend since it uses the information on the actual relative price ratio for every year in the sample. Nevertheless, it has important advantages. The problems associated with the trend rate of growth being subject to the choice of sample and ﬁxing quantity at some level are not encountered. The use of the trend rate and quantity point to hypothetical, rather than actual, cost estimates, in the sense that it computes something expressed as a counterfactual, i.e. ‘what would have happened if a particular volume of exports were exported in a certain year given the predetermined rate of falling relative

2.5

Change in relative price (%)

Relative commodity price

50 40 30 20

1.5

1980 10 0

1 −10

Change in relative price (%)

Relative commodity price: 1985=1

2

1986 −20

0.5

0

1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

−30 −40

Figure 6.1. Composite Relative Commodity Price Index and its Changes Note: The relative commodity price index is measured on the left vertical axis, while its annual percentage changes are shown on the right vertical axis.

166

Estimating Foreign Exchange Loss prices’. By contrast, the current method is based on actual export earnings and uses information on actual prices. Besides, the secular decline in relative commodity prices should be reﬂected in sustained foreign exchange losses by countries. If the problem is one of pure volatility, and not falling trend rate, one might not expect to ﬁnd any losses. By considering the possibility of encountering terms of trade gains, the proposed methodology actually does not exaggerate the foreign exchange loss over a period of time. The following provides two sets of estimates. The ﬁrst set of results is the most comprehensive, calculating the foreign exchange losses during 1995–2000 for primary commodity exports from countries which are LDCs, SVs, or HIPCs. In the second exercise, we choose a few commodities, viz. cocoa, coffee, cotton, coconut oil, palm oil, tea, jute and natural rubber and calculate the cumulative losses over 1985–2000 from exports of these commodities accruing to a number of important suppliers in our sample.

6.2. Results 6.2.1. Foreign exchange losses from primary commodity exports Falling real commodity prices are reﬂected in lower purchasing power of primary exports in that a larger volume of exports would be required in order to ﬁnance a given volume of imports. It has been estimated that between 1985 and 2000, while the volume of primary exports of all LDCs, HIPCs and SVs increased by about 55 per cent, their purchasing power actually rose by a mere 10 per cent (Figure 6.2).9 For the group of LDCs, the volume index of exports grew from 100 in 1985 to 115 in 2000, but the purchasing power index declined by about 25 percentage points. Over the same period, the purchasing power of HIPC exports increased by only about 6 per cent despite a rise of about 56 per cent in the volume index. As shown in Figure 6.2, only in the case of SVs is there a clear upward trend in the purchasing power index, which rose from 100 in 1985 to about 133 in 2000.10 Nevertheless, the trend growth rate of volume for SVS has been signiﬁcantly higher (increasing from 100 in 1985 to 148 in 2000), resulting in net foreign exchange loss. The difference between volume and purchasing power is the foreign exchange forgone out of commodity exports. Table 6.1 provides the estimated foreign exchange losses by individual LDCs, SVs, and HIPCs during the sample 9 The ‘volume’ is deﬁned as the value of exports divided by the unit value of exports. From Figure 6.1 it is observed that compared to 1985 the purchasing power of primary exports for all countries in 2000 is about 10 per cent higher. But in comparison with 1980, the purchasing power in 2000 is actually 12 per cent lower. 10 However, compared to 1980 the growth of purchasing power of primary exports of SVs has been only modest. The purchasing power index for 2002 was just about 5 percentage points higher than that of 1980 (see Figure 6.2 for SVs).

167

The Implications of Declining Commodity Prices 180

140

All Countries

160

Index: 1985=100

100

1999

2000 2000

160

Purchasing power 1998

1985

2000

1999

1998

0

HIPCs

SVs

140

60

80 60 40

40

Volume Purchasing Power 1990

0

1985

2000

20 1999

1998

1997

1996

1990

1985

1980

1995

Volume Purchasing Power

20

1998

80

100

1997

100

120

1996

120

1995

Index: 1985=100

140 Index: 1985=100

Volume

20

160

0

40

1999

180

1997

1995

1990

1985

1980

20

1996

Volume Purchasing Power

1980

40

60

1997

60

80

1996

80

1995

100

1990

120

1980

Index: 1985=100

140

0

LDCs

120

Figure 6.2. Volume and Purchasing Power of Exports Source: Authors’ estimates.

period.11 Relative to average 1984–86 base year prices, the cumulative net loss over 1995–2000 for all countries amounted to about US$37 billon, i.e. an average of just over US$6 billion per year.12 For LDCs the average annual loss is estimated to be US$2.3 billion, while the comparable ﬁgures for SVs and HIPCs are, respectively, US$0.6 billion and US$5.5 billion.13 Of the 78 countries for which estimates have been undertaken and presented in Table 6.1, as many as 67 individual countries (86 per cent of the total) are found to have suffered net

11

The choice of the sample period is determined by data availability. Note that this is the net loss incurred by the countries listed in Table 6.1. Some countries have actually experienced positive gains, and they are also taken into account in estimating the net loss (i.e. the net loss is summed over positive and negative values). 13 The annual average loss estimated for LDCs is almost the same as the UNCTAD (2002a) estimate of US$2.4 billion for the year 1998–99, using 1986 as the base year. Had the computation been carried out with 1980 as the reference year, the annual average loss by LDCs over 1995–2000 would have been much higher—about US$6 billion a year. 12

168

Estimating Foreign Exchange Loss Table 6.1. Estimated Foreign Exchange Loss by Individual LDCs, SVs, and HIPCs (US$ million in 1984–86 prices)

Country

1995

1996

1997

1998

1999

2000

Afghanistan 17.9 22.1 18.8 13.1 14.8 16.4 Antigua 0.1 0.1 0.0 0.1 0.2 0.3 and Barbuda Angola 9.6 8.3 10.9 20.7 1.7 11.3 Bangladesh 45.5 39.6 43.2 46.2 25.8 35.7 Barbados 10.4 15.1 14.2 12.7 5.4 6.4 Belize 3.6 3.8 2.4 3.5 14.0 16.5 Benin 17.7 2.8 4.6 25.6 0.4 5.4 Bhutan 5.5 9.1 11.1 12.9 14.2 13.4 Bolivia 432.7 374.3 403.0 410.0 403.2 423.5 Botswana 10.5 9.5 35.6 68.1 95.0 85.9 Burkina Faso 37.8 25.8 34.5 60.3 28.9 14.3 Burundi 57.6 51.4 130.4 114.1 157.8 160.3 Cambodia 64.3 75.7 80.2 68.3 68.5 63.1 Cameroon 62.0 37.0 14.4 32.3 227.7 174.7 Cape Verde 0.4 0.3 0.5 0.4 0.4 0.8 Central African 25.8 31.3 54.6 114.9 132.3 179.6 Republic Chad 16.1 13.0 16.1 25.1 19.7 26.0 Comoros 2.5 1.9 2.4 1.7 1.7 1.5 Congo 28.0 7.5 35.0 82.5 31.8 14.4 ˆ te d’Ivoire Co 772.5 991.0 1020.5 1044.3 949.0 838.5 Dem. Rep. 9.3 41.6 57.5 128.9 59.4 49.7 Congo Djibouti 0.5 0.7 0.6 1.5 7.7 3.3 Dominica 0.7 0.7 0.5 0.6 2.0 2.2 Equatorial 2.4 3.7 4.1 3.9 5.9 7.1 Guinea Eritrea 0.4 0.8 0.4 0.6 1.6 0.8 Ethiopia 46.4 62.5 70.0 119.1 156.3 61.7 Fiji 4.9 14.1 22.9 16.7 13.8 11.1 Gabon 142.8 86.9 113.7 376.8 399.9 429.0 Gambia 1.3 1.2 1.0 0.9 2.9 2.6 Ghana 126.2 190.7 154.0 41.3 147.4 224.0 Grenada 0.5 0.5 0.3 0.5 2.2 2.4 Guinea 29.4 29.6 30.0 26.9 25.9 29.9 Guinea-Bissau 18.2 20.2 17.7 25.9 47.0 71.8 Guyana 0.9 9.1 6.4 0.2 36.3 40.0 Haiti 23.6 22.8 23.3 21.9 17.9 19.5 Honduras 58.1 108.8 24.2 45.0 4.1 50.9 Jamaica 28.2 28.1 19.4 25.2 87.5 101.2 Kenya 323.6 397.7 429.9 422.9 226.6 236.0 Kiribati 0.1 0.3 0.6 0.5 0.5 0.4 Lao PDR 42.1 57.3 43.7 42.0 50.4 49.9 Lesotho 0.0 0.0 0.1 0.2 1.1 1.4 Liberia 1.9 3.8 4.5 5.2 11.9 32.4 Madagascar 51.0 29.6 30.3 32.8 32.9 25.1 Malawi 115.4 104.0 103.3 171.5 116.9 134.2 Maldives 5.9 4.9 8.7 8.7 6.3 7.4 Mali 13.5 14.9 26.9 15.6 52.1 50.5 Mauritania 41.5 19.6 45.5 45.7 48.0 57.4

Average Cumulative loss losses (1995–2000) (1995–2000) 103.0 0.2

17.2 0.0

39.9 164.6 64.1 17.2 45.8 66.2 2446.7 304.5 201.6 671.5 420.1 46.0 2.8 538.4

6.7 27.4 10.7 2.9 7.6 11.0 407.8 50.8 33.6 111.9 70.0 7.7 0.5 89.7

116.0 11.6 170.4 5615.7 346.3

19.3 1.9 28.4 935.9 57.7

7.8 1.8 27.0

1.3 0.3 4.5

3.1 392.6 73.8 1549.0 10.0 883.5 2.8 171.6 200.8 61.9 129.0 152.8 87.9 2036.7 2.2 285.4 2.2 59.7 201.7 745.3 42.0 173.5 257.7

0.5 65.4 12.3 258.2 1.7 147.2 0.5 28.6 33.5 10.3 21.5 25.5 14.6 339.4 0.4 47.6 0.4 9.9 33.6 124.2 7.0 28.9 43.0 (Continued )

169

The Implications of Declining Commodity Prices Table 6.1. (Continued ) Cumulative Average losses loss (1995–2000) (1995–2000)

Country

1995

1996

1997

1998

1999

2000

Mauritius Mozambique Myanmar Nepal Nicaragua Niger Papua New Guinea Rwanda Samoa Sao Tome and Principe Senegal Seychelles Sierra Leone Solomon Islands Somalia St Kitts and Nevis St Lucia St Vincent Sudan Suriname Swaziland Togo Tonga Trinidad and Tobago Tuvalu Uganda United Rep. of Tanzania Vanuatu Vietnam Yemen Zambia

2.8 41.8 225.6 17.4 45.2 91.1 25.1

7.2 82.0 317.1 31.1 100.3 110.6 63.1

7.0 102.2 407.7 89.9 126.7 119.3 110.9

23.2 86.4 483.9 51.9 85.2 158.8 84.5

5.3 89.4 535.5 65.7 132.0 168.5 66.3

10.0 74.2 520.5 49.6 157.9 196.1 53.2

35.4 476.0 2490.4 305.5 647.3 844.4 352.8

5.9 79.3 415.1 50.9 107.9 140.7 58.8

13.5 0.1 0.6

8.8 0.4 0.5

31.7 1.3 0.7

21.3 1.4 0.8

10.1 1.2 1.4

14.5 0.7 2.2

99.8 4.9 6.3

16.6 0.8 1.0

16.5 3.9 32.8 3.1 10.6 0.5 1.8 1.3 55.7 53.2 26.4 49.4 0.2 38.8

11.2 4.2 13.0 8.3 12.6 0.4 1.6 1.0 84.9 45.5 46.9 68.3 0.4 13.5

7.1 11.0 12.0 19.7 11.0 0.4 0.7 0.6 75.9 40.4 55.7 57.2 0.9 20.0

5.4 13.9 9.7 12.8 7.2 0.3 1.0 1.1 113.7 44.5 73.2 37.4 0.5 67.1

24.7 17.9 11.5 6.9 13.0 1.0 3.3 3.3 133.2 61.1 63.6 29.1 0.7 19.0

66.6 21.0 9.2 5.0 19.7 1.1 3.7 3.7 55.8 65.2 63.6 39.7 0.6 23.1

120.7 71.8 88.3 49.6 74.2 0.6 1.9 3.0 407.4 309.8 329.3 281.1 3.0 135.4

20.1 12.0 14.7 8.3 12.4 0.1 0.3 0.5 67.9 51.6 54.9 46.8 0.5 22.6

0.0 250.2 201.5

0.0 437.6 230.1

0.0 439.0 157.2

0.0 418.1 274.9

0.0 515.7 204.8

0.0 734.6 229.9

0.1 2795.2 1298.2

0.0 465.9 216.4

0.3 1.0 608.7 1119.9 28.3 35.5 25.8 112.1

2.8 1873.9 53.7 144.4

2.7 2471.8 64.1 227.2

1.6 2852.8 67.5 359.9

1.1 2940.8 92.7 247.3

9.0 11867.9 341.7 1065.0

1.5 1978.0 57.0 177.5

3147.5 4723.6 1264.5 1974.5 224.6 260.9 2867.4 4363.7

5698.5 2194.4 445.2 5122.7

7017.4 2639.7 817.7 6193.5

8429.8 3030.4 1028.5 7412.8

8048.1 2957.5 1067.4 7003.6

37064.8 14061.1 3844.2 32963.7

6177.5 2343.5 640.7 5493.9

All Countries LDCs SVs HIPCs

Note: Figures with negative sign indicate terms of trade gains.

foreign exchange losses.14 Therefore, it is not possible to sustain the argument that the net result is dominated by a few countries’ relatively large losses. Almost one-third of the estimated foreign exchange loss of all countries is attributed to Vietnam. In recent times, it has become a major supplier of a number of commodities and its increased supply is thought to have contributed 14 The estimation can show foreign exchange loss if the unit value of exports for any country rises faster than that of its imports. It is important to note that the sign associated with the loss ﬁgures in Table 6.1 can be sensitive to the choice of base year. For reasons explained above, we chose the average of 1984–86 as the base year prices.

170

Estimating Foreign Exchange Loss to falling prices, thereby affecting other traditional exporters. Among the others ˆ te d’Ivoire, Bolivia, Gabon, Myanmar, Uganda, Tanzania, and Zambia are Co which have suffered large absolute losses.15 Appendix 6.1 presents the information on the estimated average (1995– 2000) foreign exchange loss as a percentage of individual countries’ average (1995–2000) primary and merchandise exports. It is found that for 10 countries (Guinea-Bissau, Vietnam, Nepal, Yemen, Bhutan, Myanmar, Burundi, Niger, Lao PDR and Cambodia), the loss is more than 50 per cent of the value of their primary exports (see Figure 6.3). For another 12 countries (Central African Republic, Uganda, Gabon, Haiti, Tanzania, Mozambique, Bolivia, Comoros, ˆ te d’Ivoire, Zambia and Nicaragua), the forgone foreign exchange Malawi, Co constitutes at least 25 per cent of their primary exports. For all countries together, the gap between the value of primary exports and its purchasing power is equivalent to almost one-quarter of the former. In terms of merchandise exports, the net loss to all countries is about 10 per cent although, as Figure 6.3 shows, for Guinea-Bissau, Myanmar, Burundi, Niger, Central African Republic, Uganda, Tanzania, Mozambique, Bolivia and Malawi the comparable ﬁgures are more than 25 per cent. 70.0 60.0

avg. 1995–2000 loss as % of avg. 1995–2000 merchandise exports avg. 1995–2000 loss as % of avg. 1995–2000 primary exports

per cent

50.0 40.0 30.0 20.0

0.0

Guinea-Bissau Vietnam Nepal Yemen Bhutan Myanmar Burundi Niger Lao PDR Cambodia Central Af. Rep. Uganda Gabon Haiti Tanzania Mozambique Bolivia Comoros Malawi Cote d'Ivoire Zambia Nicaragua Togo Chad Botswana HIPCs LDCs Total SVs All

10.0

Figure 6.3. Foreign Exchange Loss as a Percentage of Primary and Merchandise Exports Note: both the losses and export earnings are in 1984-86 prices. Source: Based on authors’ estimates as presented in Appendix 6.1.

15 If 1980 were considered to be the base year, the foreign exchange loss would have been higher; the cumulative loss over 1995–2000 for all countries in Table 6.1 would have amounted to US$61.3 billion—i.e. an average of US$10 billion per year. Over US$3 billion of the average annual loss was due to Vietnam alone.

171

The Implications of Declining Commodity Prices

6.2.2. Foreign exchange losses by commodities The same methodology can be employed to estimate the foreign exchange losses from individual commodities exported by different countries.16 Such an exercise will require commodity speciﬁc export receipts data by country. Using the published information in the FAO Commodity Yearbook, we have estimated foreign exchange losses incurred in respect of eight major primary commodities by a number of LDCs, SVs and HIPCs, which are presented in Table 6.2. Table 6.2. Cumulative Foreign Exchange Loss from some Selected Commodities, 1985–2000 Foreign Exchange Losses based on (US$ million) Supplying countries

Cameroon ˆ te d’Ivoire Co Ghana Sierra Leone Sao Tome and Principe Papua New Guinea Afghanistan Benin Burkina Faso Central African Republic Chad ˆ te d’Ivoire Co Mali Nicaragua Senegal Sudan Togo Uganda Tanzania ˆ te d’Ivoire Co Honduras Papua New Guinea Solomon Islands Bangladesh

Foreign Exchange Losses based on (US$ million)

1980 prices

1984–86 prices

Supplying countries

1980 prices

Cocoa 852 10915.7 9807.8 73.2 89.6

1048 12002 5189.1 51.1 67.8

Bangladesh Malawi Tanzania Vietnam

Tea 367 232.7 107.6 127.8

676.5

562.3

Cotton 74.9 1017 248 77.9

18.5 291.7 132 26

1984–86 prices

121.9 342.4 178.1 256.6

ˆ te d’Ivoire Co

Coffee 1561.2

3161.7

Honduras Kenya Uganda Vietnam

2394.9 953.6 3784.4 3309.2

2035.1 1065.7 2078.2 4146.1

419.2 540.7 432.6 196.6 103 822 262.5 108.1 527.5 Palm Oil 230.5 63.3 660.7

282.9 302.3 204.3 28.9 41.5 142.9 129 8.6 151

ˆ te d’Ivoire Co Dominica Fiji Papua New Guinea Samoa St Lucia Solomon Islands Tonga Vietnam

327.8 26.8 542.2

ˆ te d’Ivoire Co Cambodia Liberia

112.5

72.3

Jute 615.1

554.5

Papua New Guinea Vietnam

Coconut Oil 40.4 111.1 9.1 4.2 32.1 30 149.4 191.7 22.6 21 21.8 12.5 5.7 5.3 5.1 4.7 38.9 46.7 Natural Rubber 480.5 143.5 281.5 87.2 389.5 51.2 35.2

8.3

1185.7

473.2

Source: Authors’ estimates based on data on exports by countries from FAO Commodity Yearbook (various issues). 16

In this case, Vx in equations (1) and (2) will represent export revenues earned from one particular commodity in question.

172

Estimating Foreign Exchange Loss The tabulated results are cumulative over the period 1985–2000, and there are actually two sets of estimates—one using 1984–86 as the base year prices, while the other takes 1980 as the reference year. As expected, in most cases, loss estimates with 1980 prices are higher than those with 1984–86 prices.17 It is now found that the persistence of price weaknesses in cocoa, cotton, palm oil, coffee, coconut oil and natural rubber resulted in very large foreign ˆ te d’Ivoire.18 The losses from cocoa for Ghana, from exchange losses to Co coffee for Honduras, Kenya, Uganda, and Vietnam, and from natural rubber for Vietnam are also very large. For cotton producing African countries, viz. Benin, Central African Republic, Chad, Mali, Sudan, and Togo, the magnitude of loss estimates are greatly inﬂuenced by the choice of base year; the use of 1980 prices, particularly, yields severe foreign exchange losses. Small island states such as Dominica, Fiji, Papua New Guinea, Samoa, Solomon Islands, St Lucia, and Tonga, have suffered from the prolonged real price declines in edible oils.

6.3. Conclusion The sustained weakness in relative prices of primary commodities has resulted in signiﬁcant foreign exchange losses to LDCs, HIPCs, and SVs. Falling real prices of commodities have caused lower purchasing power of primary exports, on which most of these countries predominantly rely for ﬁnancing their imports. The resulting foreign exchange losses relative to the total primary and merchandise exports of many of these countries are quite substantial.

17 In the case of tea and coffee, estimated losses are higher, with 1984–86 prices for most countries. 18 In interpreting foreign exchange losses, especially for individual commodities, it is important to note the ‘fallacy of composition’ argument. This view asserts that prolonged price weaknesses were caused, in part, by the expansion of commodity production in most low-income countries during the 1980s (Sapsford and Singer, 1998; Yabuki and Akiyama, 1996).

173

The Implications of Declining Commodity Prices Appendix 6.1. Estimated Average (1995–2000) Foreign Exchange Loss as Percentage of Average (1995–2000) Primary and Merchandise Exports

Country Afghanistan Angola Antigua and Barbuda Bangladesh Barbados Belize Benin Bhutan Bolivia Botswana Burkina Faso Burundi Cambodia Cameroon Cape Verde Central African Republic Chad Comoros Congo ˆ te d’Ivoire Co Dem. Rep. Congo Djibouti Dominica Equatorial Guinea Eritrea Ethiopia Fiji Gabon Gambia Ghana Grenada Guinea GuineaBissau Guyana Haiti Honduras Jamaica Kenya Kiribati

Avg. 1995–2000 Avg. 1995–2000 loss as % of avg. loss as % of avg. 1995–2000 1995–2000 primary merchandise exports Country exports 21.1 20.7 2.0

12.8 0.2 0.1

4.7 20.2 2.3 7.0 56.2 41.7 23.2 34.1 55.8 52.2 0.7 20.9

0.6 6.4 2.0 2.8 8.7 28.1 2.8 19.9 51.9 6.9 0.4 5.7

49.1

26.8

23.3 39.5

13.0 19.1

20.8 32.3 19.5

1.7 20.4 15.1

12.8

5.3

1.4 11.8

0.6 1.5

13.1 14.9 7.0 46.9 11.8 17.8 2.6 11.8 61.5

2.2 13.7 3.0 7.3 10.6 10.2 1.7 6.8 45.1

3.7 46.2 3.9 1.7 27.0 7.2

2.4 10.2 1.9 1.3 17.6 6.1

Lao PDR Lesotho Liberia Madagascar Malawi Maldives Mali Mauritania Mauritius Mozambique Myanmar Nepal Nicaragua Niger Papua New Guinea Rwanda

Samoa Sao Tome and Principe Senegal Seychelles Sierra Leone Solomon Islands Somalia St Kitts and Nevis St Lucia St Vincent Sudan Suriname Swaziland Tanzania Togo Tonga Trinidad and Tobago Tuvalu Uganda Vanuatu Vietnam Yemen Zambia

Avg. 1995–2000 loss as % of avg. 1995–2000 primary exports 52.2 6.4 14.8

12.9 0.3 2.2

20.5 35.6 20.6 11.0 17.0 2.0 43.2 56.0 57.7 25.8 54.1 6.9

17.0 28.9 14.4 5.9 9.8 0.5 25.8 35.2 8.9 19.7 39.5 3.9

33.2

25.9

8.6 21.0

7.9 15.9

5.8 21.1 42.3

2.7 12.9 18.7

7.5

7.3

18.9 0.9

11.3 0.5

0.8 1.5 14.8 12.1 20.6 44.6 24.0 7.2 15.6

0.5 1.2 9.1 10.8 11.6 31.7 12.5 6.1 1.3

7.0 47.9 8.3 58.7 56.8 28.3

3.0 41.8 7.3 18.2 2.0 19.8

Note: (1) Primary exports exclude fuels. (2) Negative ﬁgures imply terms of trade gains. Source: Authors’ estimates.

174

Avg. 1995–2000 loss as % of avg. 1995–2000 merchandise exports

7 Marginalization of LDCs and Small Vulnerable States in World Trade Bijit Bora, Roman Grynberg, and Mohammad A. Razzaque

7.1. Introduction Despite increasing global integration in terms of an unprecedented rise in the volume of trade and capital ﬂows and signiﬁcant reduction in barriers to international trade, there are indisputable concerns that some countries have failed to derive beneﬁts from the ongoing process of trade liberalization and globalization. This proposition is usually made with reference to the leastdeveloped countries (LDCs) and small vulnerable states (SVs).1 It is now becoming increasingly evident that even in this era of globalization these countries have not been able to participate effectively in global trade and commercial activities and consequently have failed to prevent their declining relative importance in world trade. Both groups of countries confront a number of problems that constrain their economic development.2 It is also generally recognized that these two groups of countries are in need of the highest

1 This paper uses the December 2001 list of 49 least developed countries as deﬁned by the Economic and Social Commission of the UN. The deﬁnition of small state covers all countries with a population of less than 1.5 million but also includes Botswana, Jamaica, Mauritius and Papua New Guinea, which are on the World Bank and Commonwealth Secretariat list of small states even though they have populations above the threshold. Appendix 1 and 2 list the countries included in LDCs and SVs. 2 The most important problems associated with small states are: (1) small size of the domestic market does not allow exploitation of increasing returns to scale in production, (2) small size prevents them from diversifying into a wide range of activities, (3) extremely narrow export base and excessive dependence on foreign trade make them vulnerable to terms of trade shocks and export instability, (4) unfavourable geographical locations make these countries isolated from world economic activities, and (5) many small states are located in regions prone to natural disasters such as cyclones and volcanic activity (Commonwealth Secretariat 1997).

175

The Implications of Declining Commodity Prices degree of consideration from the international community in support of their development efforts. A principal theme throughout the extensive research literature on the international integration problems of LDCs and SVs is ‘marginalization’. In our view, however, the notion of marginalization remains unclear. It is usually associated with not realizing the perceived gains from multilateral trade liberalization. The present chapter deﬁnes ‘marginalization’ as the declining relative importance in world trade and uses data and statistical tools to investigate the long-term marginalization trends of these two groups of countries. This paper makes an attempt to analyse the long-term declining share of LDCs and SVs in world merchandise exports and goes on to argue that the process is mostly the result of the failure of LDCs and SVs to diversify from their static comparative advantage related to the production of primary products, the signiﬁcance of which in world trade has declined considerably during the past decades. This, therefore, emphasizes the need for diversiﬁcation of exports and especially the expansion of a manufacturing export base in these countries. The chapter is organized as follows. After this introduction, section 7.3 uses historical time series data to demonstrate the decline in importance of LDCs and small states in world merchandise and commercial services export trade, while section 7.4 summarizes the long-term and recent performance of individual countries.3 Section 7.5 is an empirical analysis of marginalization of LDCs and SVs in world merchandise exports followed by some discussions on a number of factors that aggravate the process of the declining relative signiﬁcance of these two groups of countries; some concluding observations are presented in the ﬁnal section.

7.2. LDCs and Small States in World Trade: Volume, Growth Rate, and Share 7.2.1. Merchandise exports During the last 50 years world exports of merchandise goods grew by more than 100 times: from about $62 billion in 1950 to $6327 billion in 2000 (Table 7.1); developed countries posted a 106-fold increase over 1950 and developing economies fared even better registering a rise of 112 times. By these standards, the trade performances of LDCs and small states had been very modest as their exports grew by 19 and 46 times respectively, but still minimal when compared to the performance of developed and developing countries. It should also be noted that the increase was not sustained 3 Small states in this paper cover all those countries as listed in Atkins et al. (2000) and also include Jamaica, Lesotho, Namibia and Papua New Guinea but exclude Bhutan. Discussions on the deﬁnition of small states can be found in Commonwealth Secretariat (1997).

176

Table 7.1. Absolute Volume of Exports 1950

1960

1970

1975

1980

1985

1990

1992

1994

1996

1998

1999

2000

$ Billion World Developed countries Developing countries LDCs -less oilexporters Small states -less oilexporters -small-LDCs*

61.9 37.6

129.9 85.6

314.6 225.0

882.4 583.9

2022.4 1285.3

1958.6 1295.2

3478.2 2489.0

3756.1 2686.0

4278.9 2949.7

5335.3 3598.9

5454.4 3703.4

5646.7 3769.9

6326.5 3984.6

20.4

31.1

59.3

221.9

586.9

494.4

818.8

969.5

1182.9

1538.0

1549.6

1648.9

2277.3

1.8 1.6

3.1 2.7

5.2 4.5

7.2 6.0

14.6 11.4

12.0 8.8

16.4 11.4

16.2 11.7

18.1 13.5

23.5 17.3

24.4 18.9

27.6 19.5

34.4 22.0

0.6 0.32

1.2 0.65

2.2 1.35

7.4 3.5

16.4 6.5

18.2 5.6

18.7 10.2

18.7 11.4

20.9 12.9

25.9 15.3

21.7 13.3

24.9 14.2

28.4 14.4

0.04

0.07

0.09

0.16

0.29

0.25

0.32

0.39

0.44

0.52

0.53

0.50

0.47

Index: 1990 ¼ 100 World Developed countries Developing countries LDCs -less oilexporters Small states -less oilexporters -small-LDCs*

1.8 1.5

3.7 3.4

9.0 9.0

25.4 23.5

58.1 51.6

56.3 52.0

100.0 100.0

108.0 107.9

123.0 118.5

153.4 144.6

156.8 148.8

161.6 151.5

181.2 160.1

2.5

3.8

7.2

27.1

71.7

60.4

100.0

118.4

144.5

187.8

189.3

201.4

278.0

11.4 14.0

20.5 23.7

33.9 39.5

46.6 52.6

92.3 100.0

73.1 77.2

100.0 100.0

100.0 102.6

110.3 118.4

143.3 151.7

148.8 165.8

168.3 171.0

209.7 193.9

3.2 3.1

6.5 6.3

11.7 13.1

39.5 33.9

87.7 63.1

97.3 54.4

100.0 100.0

100.0 111.6

111.7 127.2

139.5 152.4

116.0 133.0

133.1 128.1

151.8 139.8

13.9

22.7

29.1

50.4

91.1

78.3

100.0

120.4

136.9

160.7

163.8

154.3

145.4

Note : Developed, Developing, and LDCs are deﬁned as in UNCTAD (2001). Lists of LDCs and small states are given in Appendix Tables 1 and 2. 46 out of a total of 49 LDCs are considered as data for Bhutan, Eritrea, and Tuvalu either are not available for the whole period of 1950–2000 or do not exist at all. Data problems also led to the exclusion of the Bahamas, Namibia, and Tuvalu from small states. There are 4 LDCs, viz., Angola, Equatorial Guinea, Sudan and Yemen, that are oil-exporting countries. Oil-exporting small states, on the other hand, are Bahrain, Equatorial Guinea, Gabon and Trinidad and Tobago. Small-LDCs are 13 small states that are also classiﬁed as LDCs viz. Cape Verde, Comoros, Djibouti, Equatorial Guinea, the Gambia, Kiribati, Lesotho, Maldives, Samoa, Sao Tome Principe, Solomon Islands, Tuvalu and Vanuatu. * indicates that the oil-exporting small-LDC, Equatorial Guinea, is not included.

The Implications of Declining Commodity Prices throughout the 50 year period. In the 1980s the combined merchandise exports of 46 LDCs and 35 small states stagnated, but increased by about $18 and $10 billion respectively in the 1990s. While it might seem encouraging that in the 1990s goods exported by LDCs more than doubled approximately 40 per cent of the $18 billion increment took place in 2000. More importantly, that surge can be attributed to four oil-exporting countries (Angola, Equatorial Guinea, Sudan, and Yemen). Between 1990 and 2000 oil-exporting LDCs increased their exports from $5 billion to $12.4 billion.4 Consequently, non-oil producing LDCs’ exports grew by just about $11 billion. Of the $6.8 billion absolute growth in 2000, almost two-thirds ($4.3 billion) is accounted for by the oil exporters while the remaining 42 non-oil exporters account for only $2.5 billion.5 The contribution of oil-exporters in the combined exports of small states is even more prominent. Bahrain, Equatorial Guinea, Gabon and Trinidad and Tobago together accounted for about 50 per cent of all small states’ exports (Table 7.1).6 Approximately 60 per cent of the absolute growth of small states’ exports of merchandise goods in the 1990s (amounting to $5.6 billion) is attributable to oil-exporters.7 It can be deduced from the constructed indices that the growth of oil-exporting countries’ manufacturing exports has been higher than that of their non-oil producing counterparts.8 The ﬁgures on merchandise imports are similar to those for exports; in the 1980s imports stagnated, while in the 1990s they increased by $17 billion and $14 billion for LDCs and small states respectively. The import indices show that the rates of expansion of merchandise imports for both groups of countries, however, have been much lower than that of the developing country group. Table 7.3 provides the information on annual average absolute growth of merchandise exports by various country groups. It shows that over the period 4 Compared to 1990 Angola’s merchandise goods export in 2000 increased by $2736 million (i.e. about 70 per cent of its 1990 exports). For Equatorial Guinea, Sudan, and Yemen the increments were respectively $335, 781, and 3541 million respectively, which are 515, 208, and 537 per cent of their respective merchandise export earnings in 1990. Between 1990 and 1998 non-oil countries’ share in all LDC merchandise exports increased from 69.4 per cent to 77.4 per cent. But mainly because of the oil export boom of Angola and Yemen the share of non-oil producing LDCs declined to 71 per cent in 1999 and then further shrank to 64 per cent in the following year. 5 Country speciﬁc merchandise export ﬁgures show that 84 per cent (i.e. $2.1 billion) of the $2.5 billion rise in non-oil countries’ exports is due to only Bangladesh and Congo. 6 Note that Equatorial Guinea is a small state as well as a LDC and, therefore, it is being considered as an oil-exporting country in both groups. 7 Among the non-oil exporters Botswana, Malta, Mauritius and Papua New Guinea together captured about three-fourths of the rise ($4.1 billion) in total merchandise exports. 8 In the 1950s non-oil exporting countries accounted for about 50 per cent of the total merchandise exports of all small states, which increased to more than 60 per cent in the early 1970s before falling to less than 40 per cent after the oil crisis of the mid-1970s. In the late 1980s to mid-1990s non-oil exporting small states’ combined share increased and hovered around 60 per cent. But just like the LDCs, during 1999–2000, oil-producing small states experienced an export boom squeezing the share of non-oil exporters to about 50 per cent in 2000.

178

Table 7.2. Absolute Volume of Merchandise Imports 1950

1960

1970

1975

1980

1985

1990

1992

1994

1996

1998

1999

2000

$ Billion World Developed countries Developing countries

63.7 41.7 18.4

136.4 88.7 34.3

327.4 236.2 61.7

902.3 612.3 205.4

2060.5 1418.1 493.6

2015.9 1387.8 467.0

3590.4 2608.7 796.2

3867.1 2746.4 1019.5

4333.2 2938.2 1251.8

5484.0 3663.1 1604.3

5556.4 3772.7 1555.9

5791.4 3970.5 1616.4

6512 4317 2112.0

LDCs -less oil-exporters

1.5 0.3

3.3 0.6

5.7 1.5

11.4 2.5

22.1 6.3

19.4 4.9

24.6 3.8

27.0 5.5

27.2 4.9

34.7 5.7

37.4 6.2

38.0 6.6

41.0 7.2

Small states -less oil-exporters -small-LDCs*

0.7 0.35 0.05

1.6 0.8 0.1

3.5 2.3 0.15

11.0 5.9 0.2

25.1 10.8 0.32

17.6 10.6 0.3

25.4 13.7 0.4

28.2 15.7 0.45

28.1 17.2 0.5

34.6 20.8 0.6

35.4 21.4 0.55

34.6 22.0 0.6

39.0 24.2 0.55

Index: 1990 ¼ 100 World Developed countries Developing countries

1.8 1.6 2.3

3.8 3.4 4.3

9.1 9.1 7.7

25.1 23.5 25.8

57.4 54.4 62.0

56.1 53.2 58.6

100.0 100.0 100.0

107.7 105.3 128.0

120.7 112.6 157.2

152.7 140.4 201.5

154.8 144.6 195.4

161.3 152.2 203.0

181.4 165.5 265.3

LDCs -less oil-exporters

6.1 7.8

13.3 15.8

23.1 39.5

46.4 65.8

89.7 165.7

79.0 128.9

100.0 100.0

110.1 144.7

110.6 128.9

141.1 150.0

152.0 163.1

154.7 173.4

166.7 189.4

Small states -less oil-exporters -small-LDCs*

2.7 2.6 12.5

6.2 5.8 25.0

13.6 16.7 37.5

43.2 43.0 50.0

99.0 78.8 80.0

69.4 77.4 75.0

100.0 100.0 100.0

111.1 114.6 112.5

110.6 125.5 125.0

136.3 151.8 150.0

139.5 156.2 137.5

136.3 160.5 150.0

153.5 176.6 137.5

Note: As for Table 1.

The Implications of Declining Commodity Prices Table 7.3. Absolute Growth of Merchandise Exports ($Billion) Average Absolute Growth Country Groups

1950–2000 1950–70 1970–2000 1970s

1980s

1990s

1990–94

1995–2000

139.92 118.7 99.93 101.6

294.8 167.5

251.46 157.23

339.67 179.04

17.12 127.3

94.23

160.63

24.25 18.61 3.64

67.06 35.44 17.62

67.22 38.15 13.7

68.74 33.18 21.38

1.24 1.12

0.51 0.52

1.55 1.12

0.82 0.41 0.018

0.66 0.67 0.03

0.87 0.06 0.07

World Developed Countries Developing Countries - HPAE -Asian Tigers -China

114.2 77.08

10.22 7.94

192.9 129

37.13

2.28

63.9

49.33

17.62 10.84 3.21

29.95 16.71 0.08

35.27 21.55 6.81

9.78 5.7 1.06

LDCs -less oil-exporters

0.43 0.34

0.13 0.11

0.61 0.47

0.65 0.58

Small States -less oil-exporters -Small LDCs*

0.55 0.31 0.016

0.07 0.04 0.009

0.75 0.49 0.03

1.14 0.21¤ 0.45 0.39 0.019 0.011

0.07¤ 0.05¤

Notes : (1) The growth of absolute exports as reported in the ﬁrst 6 columns of Table 3 is estimated using a linear trend equation: X ¼ a þ bt, where X is the exports of different country groups for different periods a is the intercept, and t is a time trend. The reported ﬁgures are estimated b’s from different regression equations. (2) All estimated coefﬁcients were statistically signiﬁcant at least at the 5 per cent level except the ones indicated by ¤. Figures in the last two columns are simple annual average absolute growth. HPAE is the group of high performing Asian economies, namely, China, Hong Kong, Indonesia, Korea, Malaysia, Singapore, Taiwan and Thailand. Asian Tigers are Hong Kong, Korea, Singapore and Taiwan. * The group of small LDCs does not include oil-exporter Equatorial Guinea.

of 1950–2000, on average, world exports increased at an annual absolute rate of $114.2 billion. The comparable ﬁgures for LDCs and small states had been only $0.43 and $0.55 billion respectively. Not surprisingly, in the 1990s when the global export volume expanded at a staggering rate of $295 billion per annum with developed and developing countries capturing respectively $168 and $127 billion, the two vulnerable groups of LDCs and small states experienced average growth rates of just over $1 billion each.9 The combined contribution of 69 LDCs and small states to annual average absolute global export growth had been only 0.71 per cent.10 Exclusion of the oil-exporters would slash this share to merely 0.31 per cent. During the last three decades (1970–2000) world merchandise exports registered a trend growth rate of 9.17 per cent as against 4.83 per cent for LDCs and 6.72 per cent for SVs. Actually, the growth rate for small states (including the oil-exporters) was higher than those of the developing and developed

9 In the case of the small states a signiﬁcant portion of the growth is due to four oilexporters alone. The last column of Table 3 shows that without the oil-exporters small states had experienced a decline in their absolute exports. 10 There are 46 LDCs for which data are available while the corresponding number of small states is 35. However, since 13 small states are also LDCs, there are a total of 69 countries in these two groups.

180

Marginalization and World Trade countries, but did not exceed the rate achieved by the high-performing Asian economies (HPAE). This impressive performance of small states was largely attributable to the oil boom of the 1970s. Small states without the oil producing countries achieved growth rates that were lower than that of world exports. In the 1980s, when oil-rich small states had experienced an absolute fall in their exports, other small states managed to enjoy a modest growth rate. Table 7.4 shows that small states without the oil-rich countries had the worst growth performance in the post-Uruguay Round period. In contrast, LDCs experienced a growth rate higher than that of world exports but lower than that of the developing country group.11 Figures 7.1 and 7.2 exposit graphically the information and data contained in Tables 3 and 4, that over time LDCs and small states have become marginalized in the world merchandise export trade, failing to protect their relative importance. The share of all LDCs in world merchandise exports has gone down from about 3.5 per cent in 1954 to 0.54 per cent in 2000. A similar trend is also observed when the oil-exporting countries are excluded—a secular decline from more than 3 per cent in 1954 to 0.35 per cent in 2000.12 Like LDCs, small states’ shares have also been subjected to a long-term declining trend, falling from 1.18 per cent in the mid-1950s to 0.44 per cent Table 7.4. Trend Growth Rates of Exports (per cent) Average Growth Rate Country/Country Groups World Developed countries Developing countries - HPAE - Asian Tigers - China

1950–2000 1950–70 1970–2000 1970s 1980s

1990s 1990–94 1995–2000

10.29 10.39

7.26 8.31

9.17 9.01

18.82 17.51

5.11 6.46

6.37 5.36

7.34 6.68

6.93 5.55

10.1

5.13

9.54

21.95

2.24

8.51

8.9

9.75

12.84 14.54 11.66

4.2 5.6 6.1

14.38 14.97 14.88

24.07 10.82 25.08 13.28 17.59 11.43

9.56 8.26 13.52

13.96 12.19 18.46

8.4 6.84 13.32

LDCs - less oil-exporters

5.19 4.91

3.99 3.85

4.83 4.52

0.54¤ 0.56¤

6.36 7.22

3.46 4.39

7.92 8.69

Small States - less oil-exporters - Small LDCs*

8.42 8.38 5.11

5.83 5.96 1.01

6.72 7.53 7.26

20.34 1.44¤ 15.85 5.33 10.58 3.76

4.14 3.29 3.35

3.52 6.18 8.51

4.03 0.001¤ 0.01¤

8.6 8.9

Note : The trend growth rates are estimated by ﬁtting a semi-logarithmic trend equation to the data. Except for the ones denoted by ¤, all the trend coefﬁcients were found to be statistically signiﬁcant. * The group of small LDCs does not include oil-exporter Equatorial Guinea.

11

Previously it was mentioned that four oil-producers and two other large exporters, Bangladesh and Congo, were mainly responsible for a better performance of the group of LDCs. 12 Due to an export boom of oil-exporters all-LDCs’ share increased from 0.45 per cent in 1998 to 0.54 per cent in 2000. But when oil-producing countries are excluded the share remained stagnant at 0.35 during the same time.

181

4.00 3.50 3.00

All LDCs

LDCs without oil-exporters

2.50 2.00 1.50 1.00 0.50 1998

1995

1992

1989

1986

1983

1980

1977

1974

1971

1968

1965

1962

1959

1956

1953

0.00 1950

share in world merchandise exports (per cent)

The Implications of Declining Commodity Prices

Figure 7.1. Share of LDC Exports in Global Merchandise Exports: 1950–2000 Note: The ﬁgure is based on 46 LDCs for which data are available. Source: Estimated from UNCTAD (2001) and WTO database.

1.4 Share of small states

Share of small states without oil-rich countries

1.2

Per cent

1 0.8 0.6 0.4 0.2

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

1977

1975

1973

1971

1969

1967

1965

1963

1961

1959

1957

1955

1953

1951

Year

0

Figure 7.2. Share of Small States in Global Merchandise Exports: 1950–2000 Source: Authors’ estimates from UNCTAD (2001) and WTO database.

in 2000. The comparable ﬁgures for small states, excluding oil dependent countries, are respectively 0.68 and 0.23. The graph representing all small states in Figure 7.2 exhibits a sudden jump in 1974 and then portrays a shortlived episode of enhanced but deteriorating share with respect to the secular

182

Marginalization and World Trade declining trend of the 1960s. This was due to the export boom of the oilproducing small states following the oil crisis of 1973. By the middle of the 1980s all small states (including the oil-exporting countries) had reversed back to the trend that was originated in the early 1960s, as the oil-exporting countries suffered steep declines in their relative importance, and this trend continued throughout the 1990s.13

7.2.2. Exports of commercial services One of the salient features of globalization has been the growing trade in commercial services that include travel, transport, communications and ﬁnancial and professional services. Technological advancement leading to introduction of new sectors and reduction of costs involving computation and communications has made many services tradable resulting in a dramatic expansion in transaction of services (Hoekman and Braga, 1997). Exports of commercial services are now equivalent to about a quarter of merchandise exports. The rising importance of the services trade resulted in the inclusion of the General Agreements of Trade in Services (GATS) in the Uruguay Round of MTNs. According to the information provided in Table 7.5 world exports of commercial services stood at $1457 billion in 2000, of which about 72 per cent accounted Table 7.5. Exports of Commercial Services ($billion) Country Groups

1980

1985

1990

1992

1994

1996

1998

World Developed Countries Developing Countries LDCs Small States - Major 4 - Small-LDCs

364.3 277.7

381.8 288.9

783.2 608.7

924.5 708.3

1039.5 760.8

1276 912.8

1341 974.2

74

80.6

153.8

190.4

245.1

310.5

314

2.4 5.1 1.5 0.2

3.3 9.3 3.9 0.4

3.6 10.8 4.7 0.5

4.3 12.3 5.4 0.6

5.6 13.9 6.2 0.7

2.3 4.3 1.3 0.2

1999 1376 1006

2000 1457.2 1046.3

323.9

360.6

5.6 14.5 6.5 0.5

5.9 15.8 7.1 0.9

6.0 16.3 7.1 0.9

Index 1990 ¼ 100 World Developed Countries Developing Countries LDCs Small States - Major 4 - Small-LDCs

46.5 45.6

48.7 47.5

100.0 100.0

118.0 116.4

132.7 125.0

162.9 150.0

171.2 160.0

175.7 165.3

186.1 171.9

48.1

52.4

100.0

123.8

159.4

201.9

204.2

210.6

234.5

69.7 46.2 33.3 50.0

72.7 54.8 38.5 50.0

100.0 100.0 100.0 100.0

109.1 116.1 120.5 125.0

130.3 132.3 138.5 150.0

169.7 149.5 159.0 175.0

169.7 155.9 166.7 125.0

178.8 169.9 182.1 225.0

181.8 175.3 182.1 225.0

Note : The major 4 includes Cyprus, Jamaica, Malta and Mauritius. Source : Authors’ estimates from WTO database. 13 During the past 50 years the share of all Commonwealth small states has also been squeezed considerably, which is shown in Appendix 3.

183

The Implications of Declining Commodity Prices for developed countries.14 LDCs posted only $6 billion worth of commercial services in 2000 and during the last 20 years their exports grew by merely less than $4 billion. In that respect small states perform much better as their earnings increased from $4.3 billion in 1980 to $16.3 billion. However, about 50 per cent of this absolute growth was due to 4 countries alone—viz. Cyprus, Jamaica, Malta, and Mauritius—that have relatively large and well-developed tourism sectors. The data on imports of commercial services, as presented in Table 7.6, show that LDCs are much more dependent on imported commercial services than small states and the rate of expansion in the import demand by the former has been higher.15 Table 7.7 provides the estimates of absolute growth of commercial services exports. Between 1980 and 2000 global services exports grew at an annual average rate of about $63 billion; developed countries experienced a per annum growth of $44.6 billion and developing countries $17.5 billion. In contrast, LDCs and small states registered annual absolute growth of $0.22 and $0.58 billion respectively. From the information provided in Table 7.7 it can be estimated that almost 60 per cent of the growth in small states was attributable to 4 major countries only. Without these four countries, the combined contribution of LDCs and small states in world growth of commercial services in the 1990s stood at only 0.79 per cent. Table 7.6. Imports of Commercial Services ($billion)

World Developed Countries Developing Countries LDCs Small States - Major 4 - Small-LDCs

1980

1985

1990

1992

1994

1996

1998

1999

2000

400.4 265.5 121.8 6.7 4.4 1.0 0.3

399.8 273.1 114.1 5.8 4.6 1.0 0.2

817.9 603.6 190.1 8.9 6.7 2.2 0.4

941.7 686.3 226.7 9.9 7.6 2.6 0.5

1043.3 734.2 273.8 9.9 7.9 2.9 0.5

1266.5 874.1 347.7 12.2 9.2 3.7 0.7

1335.5 938.1 350.0 12.9 9.9 3.8 0.8

1367.3 966.7 357.1 12.9 10.0 3.9 1.0

1453.1 1004.7 396.8 14.5 10.6 4.1 1.2

154.8 144.8 182.9 137.1 137.3 168.2 173.8

163.3 155.4 184.1 144.9 147.8 172.7 197.6

167.2 160.2 187.8 144.9 149.3 177.3 238.1

177.7 166.5 208.7 162.9 158.2 186.4 285.7

Index 1990 ¼ 100 World Developed Countries Developing Countries LDCs Small States - Major 4 - Small-LDCs

49.0 44.0 64.1 75.3 65.7 45.5 59.5

48.9 45.2 60.0 5.2 68.7 45.5 57.1

100.0 100.0 100.0 100.0 100.0 100.0 100.0

115.1 113.7 119.3 111.2 113.4 118.2 121.4

127.6 121.6 144.0 111.2 117.9 131.8 121.4

Note : The major 4 includes Cyprus, Jamaica, Malta, and Mauritius. Source : Authors’ estimates from WTO database. 14 Developed countries are the most important service suppliers in world. Having perceived this comparative advantage, the US, backed by other developed countries, sought to bring services transactions under trade rules and regulations despite severe protests by the developing and least developed countries. 15 This is reﬂected in the constructed index; the index for LDCs rises to about 209 as against 163 of small states. Small-LDCs, however, have the biggest rates of expansion in the demand for imported commercial services.

184

Marginalization and World Trade Table 7.7. Absolute Growth of Commercial Services Exports ($billion) Country groups World Developed Countries Developing Countries - HPAE - Asian Tigers - China LDCs Small States - Major 4 - Small-LDCs

1980–2000

1980s

1990s

62.9 44.6 17.5 7.7 5.7 1.4 0.22 0.58 0.34 0.04

32.4 26.3 5.7 3.1 2.4 0.27 0.07 0.38 0.22 0.012

71.1 46.0 23.9 9.3 7.0 2.5 0.32 0.64 0.34 0.061

Source : Authors’ estimates from WTO database.

The 1980–2000 trend growth rates of services exports for LDCs and small states are estimated to be respectively 5.9 and 7.4 per cent (Table 7.8), which are considerably lower than those of world, developed, and developing country groups (Table 7.8).16 Therefore, even after experiencing a steady increase in the volume of services exports, LDCs’ and small states’ relative importance in the world actually shrank. For LDCs the fall in share was from 0.6 per cent in 1980 to 0.4 per cent in 2000 and for small states the corresponding ﬁgures are respectively 1.18 and 1.12 per cent. Table 7.8 shows that the decade of the 1980s was the time when small states enjoyed a higher growth rate than all other country groups. In the following decade, however, small states were outperformed by the developing as well as least developed countries.17 Table 7.8. Growth Rates of Exports of Commercial Services Country groups World Developed Countries Developing Countries HPAE - Asian Tigers - China LDCs Small States - Major 4 - Small-LDCs

1980–2000

1980s

1990s

8.4 8.1 9.3 12.5 11.6 14.5 5.9 7.4 10.01 9.26

6.8 7.3 5.5 9.8 9.5 8.0 2.1 7.8 11.0 5.29

6.5 5.6 8.8 10.1 9.0 16.7 7.2 5.8 6.38 9.64

Note : The reported ﬁgures are trend growth rates. Source : Authors’ estimates from WTO data. 16 High performing Asian economies have growth rates even higher than the world and developing countries. 17 Figure 7.3 reveals that the relatively superior growth performance of the small states was concentrated only in the early- to mid-1980s resulting in their rising share. Growth rates in the latter half of the 1980s were actually lower than the world causing relative signiﬁcance to fall.

185

The Implications of Declining Commodity Prices Small States

1.6

LDCs

Small without big 4

1.4

Per cent

1.2 1 0.8 0.6

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

1984

1983

1982

1981

0.2

1980

0.4

Figure 7.3. Share of Small States and LDCs in Commercial Services Exports

7.2.3. Total export (merchandise plus commercial services) trade The preceding discussions have shown the diminishing importance of LDCs and small states in both merchandise and commercial exports of the world. Table 7.9 now combines both these exports explicitly to consider the relative performance of the two vulnerable groups of countries in global export trade. One can estimate from Table 7.9 that in 2000 the combined share of LDCs and small states in world export trade stood at about 1 per cent—down from 1.5 per cent of 1980.

Table 7.9. Volume of Export Trade (Merchandise Plus Commercial Services) ($ billion) 1980

1985

1990

1992

1994

1996

1998

1999

2000

World 2386.7 2340.4 4261.4 4680.6 5318.4 6611.1 6795.4 7022.5 7783.7 Developed Countries 1563.0 1584.1 3097.7 3394.3 3710.5 4511.7 4677.6 4776.3 5030.9 Developing Countries 660.9 575.0 972.6 1159.9 1428.0 1848.5 1863.6 1972.8 2637.9 LDCs 16.9 14.4 19.7 20.0 22.4 29.1 30.0 33.5 40.4 Small States 20.7 23.3 28.0 29.5 33.2 39.8 36.2 40.7 44.7 Small-LDCs 0.5 0.5 0.8 1.0 1.2 1.6 1.6 2.0 2.0 Index 1990 ¼ 100 World Developed Countries Developing Countries LDCs Small States Small-LDCs

186

56.0 50.5 68.0 85.8 73.9 62.2

54.9 51.1 59.1 73.1 83.2 57.3

100.0 100.0 100.0 100.0 100.0 100.0

109.8 109.6 119.3 101.5 105.4 123.2

124.8 119.8 146.8 113.7 118.6 148.8

155.1 145.6 190.1 147.7 142.1 192.7

159.5 151.0 191.6 152.3 129.3 195.1

164.8 154.2 202.8 170.1 145.4 241.5

182.7 162.4 271.2 205.1 159.6 245.1

Marginalization and World Trade 100.0 LDCs

95.0

Small States

Index: 1980=100

90.0 85.0 80.0 75.0 70.0 65.0 60.0

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

1984

1983

1982

1981

50.0

1980

55.0

Figure 7.4. Declining Importance of Small States and LDCs in World Export (Merchandise Plus Services) Trade

Figure 7.4 measures the fall in shares of small states and LDCs in world export trade during the last two decades. It shows that between 1980 and 2000 small states’ share has fallen by about 36 per cent while the comparable ﬁgure for LDCs is 30 per cent.18

7.2.4. Total trade transactions Total trade transactions (i.e. exports and imports of goods and commercial services taken together) of the world economy stood at $15749 billion of which two-thirds were attributable to developed countries. LDCs, with more than 13 per cent world population, account for just about 1.8 per cent of trade transactions of the developing countries. Table 7.10 shows that LDCs and small states have almost similar trends in trade growth and they have had comparable trading volumes.19 18 Until 1994 there had been a secular decline in the share of LDCs, between 1995–97 the share remained more or less unchanged; in 1999 and then in 2000 it shot up. As discussed in an earlier section, the exports boom of the oil producing countries was mainly responsible for this rise. 19 The growth of trade is reﬂected in the estimated index (1990 ¼ 100) in Table 7.10. The index for LDCs rose from 86 in 1980 to 183 in 2000 while for small states the corresponding ﬁgures were 84 and 157. Note that the rate of expansion of world trade transactions has been much faster as its index started at 56 in 1980 and reached 182 in 2000. For the developing world the growth has been even higher. Since information on commercial services is available since only 1980, the data on total trade transactions for previous decades are not given.

187

The Implications of Declining Commodity Prices Table 7.10. Total Trade Transactions of Different Country Groups ($ billion) 1980

1985

1990

1992

1994

1996

1998

1999

2000

World 4847.6 4756.1 8669.7 9489.4 10694.9 13361.6 13687.3 14181.2 15748.8 Developed 3246.6 3245.0 6310.0 6827.0 7382.9 9048.9 9388.4 9713.5 10352.6 Countries Developing 1276.3 1156.1 1958.9 2406.1 2953.6 3800.5 3769.5 3946.3 5146.7 Countries LDCs 45.7 39.6 53.2 56.9 59.5 76.0 80.3 84.4 95.9 Small States 50.2 45.5 60.1 65.3 69.2 83.6 81.5 85.3 94.3 Small-LDCs 1.96 1.81 3.04 4.12 4.13 5.41 5.43 5.68 6.21 Index 1990 ¼ 100 World Developed Countries Developing Countries LDCs Small States Small-LDCs

55.9 51.5

54.9 51.4

100.0 100.0

109.5 108.2

123.4 117.0

154.1 143.4

157.9 148.8

163.6 153.9

181.7 164.1

65.2

59.0

100.0

122.8

150.8

194.0

192.4

201.5

262.7

85.9 83.5 64.5

74.4 75.7 59.5

100.0 100.0 100.0

107.0 108.7 135.5

111.8 115.1 135.9

142.9 139.1 178.0

150.9 135.6 178.6

158.6 141.9 186.8

180.3 156.9 204.3

Finally, Figure 7.5 shows the shares of small states and LDCs in the world’s total trade transactions. It is shown that both groups of countries’ relative importance fell rapidly in the 1980s. The 1990s is shown not to cause any trend decline but our previous discussions showed that oil-exporting countries might have contributed to this apparently improved performance by these vulnerable groups of countries. 1.20 LDCs

Small States

1990

1994

1.00 0.80 0.60 0.40 0.20 0.00 1980

1985

1998

Figure 7.5. Share of Small States and LDCs in World Trade Transactions

188

2000

Marginalization and World Trade

7.3. Performance of Individual Countries 7.3.1. Long-term trends Countries included in the groups of LDCs and small states are not homogenous at all and the performance of individual countries differs wildly both in terms of absolute exports (Appendices 7.3 and 7.4) and their growth rates (Appendices 7.5 and 7.6). Also, in terms of dependence on merchandise goods in total exports earnings, these countries differ a lot. In general small states’ export baskets are dominated by services exports while LDCs mostly rely on merchandise products.20 Notwithstanding this dissimilarity, high volatility of exports stands out to be the most striking characteristic associated with these countries. Graphical plots of aggregate exports of individual LDCs and small states, as given respectively in Figures 7.6 and 7.7, show that an overwhelming majority of these countries has been subject to frequent and at times violent ﬂuctuations in their export receipts.21 Among the LDCs, only six countries, namely, Bangladesh, Benin, Bhutan, Lao PDR, Myanmar, and Nepal have experienced relatively stable export earnings. For Afghanistan, Burundi, Haiti, Niger, Rwanda, Sierra Leone, and Zambia, along with recurrent ﬂuctuations, declining trends in their aggregate export receipts are discernible.22 In the set of small states, countries that have undergone most spectacular variability in exports are the Gambia, Kiribati, Sao Tome and Principe, Suriname, and Trinidad and Tobago. The problem of volatility in export earnings arises both from merchandise goods as well as in commercial services exports (see the ﬁgures given in Appendix 7.7–7.10). Dependence on a narrow range of agricultural products that experience either price or exogenous supply shocks on a regular basis is the principal reason for export instability in merchandise exports. In the case of commercial services, most LDCs and small states rely on the travel and tourism sector, which, in turn, is dependent on such factors as the global political situation that is completely beyond the control of these countries. A close examination of Appendix Figures 7.3–7.6 reveals that for most individual small states, earning instabilities are more prominent in the case of merchandise exports, but no such conclusion can be derived for LDCs. The high degree of instability in export earnings is also reﬂected in individual countries’ volatile relative signiﬁcance in world trade, which is depicted in Figures 7.8 and 7.9. These ﬁgures sketch out the evolution of individual LDCs’ and small states’ share in total world exports by setting their actual shares in 1980 20

See last two columns in Appendix Tables 7.3 and 7.4. Aggregate exports comprise merchandise and commercial services export receipts. Since the data on services exports are available only from 1980, the ﬁgures depict the aggregate exports for the last 20 years. 22 Afghanistan, Burundi, Rwanda and Sierra Leone are also countries with civil strife. 21

189

The Implications of Declining Commodity Prices Afghanistan

750

7500

Angola

Bangladesh

750

150

Benin

6000 500

5000

500

100

2500

250

50

Bhutan

4000

250

2000 1980

400

1990

2000

1980

1990

2000

1000

Burundi

Burkina Faso

1980

150

300

1990

2000

1980

200

Cambodia

1990

2000

Central African Rep.

150

200

100

100

50 1990

2000

1990

2000

800

1980

1990

2000

1000

1000

Congo, Dem. Rep. of

3000

1980

Ethiopia

1980

1990

2000

1990

2000

600

200

1990

2000

1990

2000

600

600

Mauritania

500

1980

1990

2000

500

1990

2000

Malawi

1500

400 300 1990

2000

1980

1990

2000

100 1990

2000

1990

2000

1990

2000

750

1990

2000

400

1980

1990

2000

2000

1000

1990

2000

1990

2000

1980

1990

2000

Zambia

1000

2000 1980

2000

Sudan

Yemen, Republic of 1500

3000

250

1990

500 1980

4000

Uganda

1980

Somalia

100 1990

300 1980

Niger

500

150

500

200

1980

600

N epal

200

Sierra Leone

1980

Togo

400

500

2000

100 1980

500

Tanzania

1990

300 1980

150

1500 1000

1000

200

750

50

2000

200 1980

500

250

1000

1980

Myanmar

Senegal 1250

150

1990

Mali

400

500

Rwanda 200

2000

1000

200 1980

1990

1980

600

400

1980

2000

Mozambique

400

2000

100 1980

300 1980

1990

Haiti

200

600

Madagascar

400

400

2000

150

500

500

1990

Chad

1980

250 Guinea-Bissau

25 1980

600

Liberia

Lao PDR

400

1980

2000

50

500

400

1990

75

600 1000

2000

100 1980

Guinea 750

2000

300

1990

200

500 100

1980

1980

1000 1980

1990

2000

1980

1990

2000

1980

1990

2000

Figure 7.6. Aggregate Exports (Merchandise Plus Services) of Individual LDCs ($million) Note : Aggregate exports of the least developed countries that are also small states are shown in Figure 7.7.

to 1. A value greater than 1 in any of the following years will indicate a rise in the relative importance of the country in question in world trade and vice versa. Figure 7.8 shows that despite very wide ﬂuctuations there are as many as 24 LDCs (out of a total of 35), viz., Afghanistan, Angola, Burundi, Central African Republic, Chad, Congo, D.R., Ethiopia, Guinea, Haiti, Liberia, Madagascar, Malawi, Mauritania, Mozambique, Niger, Rwanda, Senegal, Sierra Leone, Somalia, Sudan, Tanzania, Togo, Uganda, and Zambia with declining 190

Marginalization and World Trade Antigua and Barbuda 400 200

1980

100

1990

7000 6000 5000 4000 3000

2000

2000

600

1990

2000

1990

Fiji

4000

1990

2000

1980

1990

2000

3000

500

Jamaica

1990

2000

2000

Gambia

1990

2000

3000

Malta

1000 1990

2000

1980

250

Seychelles

200

1990

2000

Solomon Islands

1990

2000

1990

2000

Suriname

1000

1990

2000

Swaziland

500

500

1990

2000

Samoa

1990

2000

St Kitts and Nevis

1990

2000

St Lucia

1990

2000

1990

2000

Trinidad and Tobago

1990

2000

St Vincent and the Grenadines

1980

150

4000

1990

2000

Vanuatu

100

3000

20 1980

2000

50 1980

5000

Tonga

1990

100

100 2000

Sao Tome

1980

150

300

1990

2000

10 1980

400

1990

20

50

2000 1980

Maldives

1980

30

25

1980

30

2000

200 1980

75

50 1980

1990

300

50

1980

150

1980

400

Lesotho

200

100 1980

PN Guinea

100

150

200

2000

1000

1000 1980

1990

2000

2000

Grenada

100 1980

3000

Mauritius

2000

50 1980

300

1990

100

50 1980

2000

Dominica

1980

150

100

100 1990

2000

1000

150

Gabon

1990

200

1980

2000

50 1980

10 1980

1990

100

20

250

400

2000

30

2000

1980

150

Djibouti

30 1990

Kiribati

Guyana

2000

40

2000 1980

Botswana

1000 1990

50

3000

200

3000 2000

1980

60

Cyprus

1000 2000 1980

1000

3000

2000

2000

400

750

1990

3000

1980

1500

Eq Guinea

Belize

200

1980

4000

Comoros

20 1980

300

100 1990

40 50

Barbados

1000

1980

60

Cape Verde

1500

Bahrain

1980

1990

2000

1980

1990

2000

1980

1990

2000

Figure 7.7. Aggregate Exports (Merchandise Plus Commercial Services) of Individual Small States Note : Exports are in millions of US dollars.

relative importance, while there are only four countries, viz. Bangladesh, Bhutan, Cambodia, and Nepal that have shown some clear rise in their shares over the long-run. Similarly, it is obvious from Figure 7.9 that small states, in general, have not been able to improve their share on a long-term basis. For Bahrain, Barbados, Belize, Djibouti, Fiji, Gabon, Gambia, Guyana, Jamaica, Sao 191

The Implications of Declining Commodity Prices 1.25

Afghanistan

Angola

Bangladesh

1

2

2

1.5

1.5

1 .5

.75 1990

2000

1980

Burkina Faso

1.5

1990

2000

Burundi

1980

15

2 1

Bhutan 3 2

1

1 1980

Benin

1990

2000

1

1980

1990

2000

1980

1990

2000

1980

1

2000

Central African Rep

.5

1980

Ethiopia

1 Congo Dem. Rep. of .75

5 4 3 2 1

1990

2000

1980

1

Lao PDR

2000

2000

Guinea-Bissau

1.5

1

1990

2000

1990

2000

1990

2000

Haiti

.5

1980

1990

2000

Madagascar

1980

1990

2000

1980

Malawi 1

1

.75 .75

.5

1980

1

1

Liberia

.75

1990

2

.75 1990

2

1

1980

2.5

1

.5 1980

2000

1.25

.75

.5

1990

Guinea

1

2000

1.5

5 1990

1990

Chad

Cambodia

10 1

.5

1980

Mali

.8

.5 .5

1980

2

1990

2000

1980

1

Mauritania

1.5

1990

2000

1980

1990

2000

1980

Mozambique Myanmar

1

.75

2

1990

2000

1

Nepal

.5

1980

1990

2000

1.5

Rwanda

1.5

1980

1

.5 1990

2000

1

1980

1

Senegal

1

1990

2000

Sierra Leone

1990

2000

1

Tanzania

1

1980

.5

1990

2000

1980

1990

2000

.5

2000

1980

1990

2000

Sudan

Somalia

1 .5

1980

Uganda

1990

2000

Yemen, Republic of

Togo

1980

1

1990

2000

Zambia

1.5 .5

1

.5

.5

1990

.5

1

.75

1980

1

.5 1980

Niger

.5

1.5 1

1980

.5 1980

1990

2000

1980

1990

2000

1980

1990

2000

1980

1990

2000

1980

Figure 7.8. Share of Individual LDCs in World Aggregate Exports, 1980–2000 Note : Share of individual small states in 1980 has been set to 1. Source : Authors’ own estimates.

192

1990

2000

Marginalization and World Trade

3

Bahrain

Antigua and Barbuda

1.5

1

2 .5

1 1980

3 2.5 2 1.5 1

2000

1990

1980

1990

2000

Fiji

1

2000

1 1980

1.5

1990

2000

Jamaica

1990

2000

Gabon

2000

Dominica 3 2

1980

1990

2000

1

1990

2000

Kiribati

1990

2000

Grenada

1.5 1.25

.5 1980

1980

1.75

Gambia

1.5

1 1980

1.25

.5

1990

2000

1980

2.5

Lesotho

1

2

.75

1.5

.5 1990

2000

1980

1990

2000

1980

1990

2000

1990

2000

Maldives

1 1980

1.2

PN Guinea

Mauritius

Malta

1990

2000

1980

1990

Seychelles

2000

1.25 1

1

.75 1990

2000

1990

2000

Solomon Islands

.5

1980

1.75

1990

2000

1980

1990

2000

2.5

1.5

1 .5

1980

1990

2000

1

.8

.75

.6

.5 1980

1990

2000

St. Vincent & Grenadines

1 1980

1990

2000

1980

1990

2000

Tonga

1980 1.25

Trinidad and Tobago

1 Swaziland

2000

1.5

1 2000

1990

2

1.25 1.25

1990

1980

St. Lucia

St. Kitts and Nevis

1.5

1980

Suriname

Sao Tome

.75

1 1980

2000

.8 1980

1.2

1990

1 1

1

.75

1980

1

Samoa

1.25

1.5

1

1

1990

.8 1980

1.4

1980

1

1

.9

.5

1.25

2000

Djibouti

1 .75

1990

1

.5 1980

Guyana

2000

.75

.6 1990

1990

Cyprus

1980

1

.8

1980

2

.8 1980

Botswana

3

1

Eq Guinea

5

.9

1.5

1 2000

1

2

Comoros

1.5

1990

1

2000

Belize

1.1

.75 1980

2

Cape Verde

1980

10

1990

Barbados

1.25

1990

2000

Vanuatu

1 .5 .75 1980

1990

2000 1980

1990

2000

1980

1990

2000

Figure 7.9. Share of Individual Small States in Aggregate Global Exports, 1980–2000 Note : Share of individual small states in 1980 has been set to 1. Source : Authors’ own estimates.

193

The Implications of Declining Commodity Prices Tome and Principe, Solomon Islands, Suriname, Tonga, Trinidad and Tobago, and Vanuatu a clear deteriorating trend in their share in world exports is found. Appendix Figures 7.11–7.14 give similar information on shares of individual countries disaggregated by merchandise goods and commercial services exports, where it is found that in both types of exports most LDCs and small states have very volatile and declining relative signiﬁcance. A country can only prevent marginalization (or, relative importance for that matter) if its exports grow at a rate at least as high as that of the world average. The above graphical analysis, therefore, implies that for most LDCs and small states the trend growth rate exports have been lower than that of the rest of the world. Tables 7.11 and 7.12 present the estimated trend growth rates of Table 7.11. Growth Rates of Merchandise and Services Exports from Individual LDCs Merchandise Exports Countries World Afghanistan Angola Bangladesh Benin Bhutan Burkina Faso Burundi Cambodia Central African Rep. Chad Congo, D.R. Ethiopia Guinea Guinea-Bissau Haiti Lao PDR Liberia Madagascar Malawi Mali Mauritania Mozambique Myanmar Nepal Niger Rwanda Senegal Sierra Leone Somalia Sudan Tanzania Togo Uganda Yemen, Republic of Zambia

Commercial Services Exports

1970–2000

1980s

1990s

1980–2000

1980s

1990s

9.17 0.6 7.3 9.9 11.0 11.4* 8.0 2.9 17.0 4.9 6.8 2.2 2.6 8.5 7.9 2.4 10.1 1.1 0.9 6.0 3.1 5.3 0.1 6.8 8.4 5.6 1.7 5.1 6.4 3.4 1.5 1.9 5.8 10.1 10.1 0.6

5.11 7.4 3.0 6.4 9.7 18.3 6.7 4.7 1.4 0.5 8.0 4.4 0.3 2.2 3.0 1.1 10.5 1.7 1.4 0.1 3.4 5.6 9.0 11.0 7.6 6.6 0.1 3.0 3.3 3.4 1.3 6.4 0.2 0.7 1.5 0.5

6.37 10.0 2.7 14.4 3.1 7.5 12.4 4.4 19.1 8.2 2.5 8.1 11.0 4.7 14.5 3.9 14.1 5.4 2.8 2.0 5.6 1.1 7.2 12.7 10.2 0.6 3.6 3.8 34.9 1.0 11.0 7.1 7.5 14.9 18.9 2.9

8.4 5.5 1.1 3.4 7.1 — 7.4 0.7 12.9 0.2 10.1 1.1 6.5 1.7 — — 16.1 6.7 11.1 0.7 4.1 0.2 7.5 13.6 10.2 5.7 1.1 3.3 8.3 — 13.7 9.9 0.6 — — 0.5

6.8 20.6 0.7 4.1 6.1 — 6.3 0.3 9.4 7.5 30.5 12.7 11.6 5.6 — — 3.1 12.9 9.8 1.5 9.3 2.6 4.0 0.7 3.9 3.8 7.8 3.7 8.1 — 10.3 4.2 3.6 — — 10.5

6.5 7.9 6.7 3.8 2.3 — 2.3 15.0 14.3 5.7 5.0 2.5 5.5 0.3 1.5 — 20.7 2.3 9.9 1.5 0.5 8.3 10.3 22.4 11.1 5.8 3.7 1.2 7.4 — 11.7 16.0 5.0 22.7 5.7 0.3

Note : * implies average growth rate of period 1980–2000.—suggests that data are not available.

194

Marginalization and World Trade Table 7.12. Growth Rates of Merchandise and Services Exports from Individual Small States Merchandise Exports Country World Antigua and Barbuda Bahrain Barbados Belize Botswana Cape Verde Comoros Cyprus Djibouti Dominica Equatorial Guinea Fiji Gabon Gambia Grenada Guyana Jamaica Kiribati Lesotho Maldives Malta Mauritius Papua New Guinea Samoa Sao Tome and Principe Seychelles Solomon Islands St Kitts and Nevis St Lucia St Vincent and the Grenadines Suriname Swaziland Tonga Trinidad and Tobago Vanuatu

Commercial Services

1970–2000

1980s

1990s

1980–2000

1980s

1990s

9.17 3.92 7.49 5.32 5.92 14.28 6.89 2.43 7.89 0.40 8.25 10.57 5.92 7.53 1.41 4.32 3.48 3.94 4.38 11.44 12.68 12.09 10.05 8.08 1.20 0.97 13.05 8.71 2.22 8.18 9.78 3.39 8.38 4.71 3.10 1.35

5.11 6.6 5.9 3.2 1.7 18.6 6.8 2.8 3.5 8.7 15.7 17.9 0.5 6.7 1.4 7.9 4.2 1.0 2.3 2.2 20.8 6.8 13.0 5.4 1.6 4.2 6.3 1.2 3.6 9.9 16.3 1.6 2.9 0.5 12.6 6.7

6.37 0.82 2.9 3.9 6.05 3.8 10.8 9.0 1.4 0.7 0.5 24.4 3.0 5.4 16.9 0.5 8.7 2.1 6.2 10.8 5.6 5.9 3.5 3.0 11.3 7.4 14.1 4.5 0.6 8.8 5.4 0.3 5.7 0.8 6.4 4.7

8.4 11.3 2.37 6.32 13.0 8.34 8.76 19.4 11.7 3.86 15.7 1.72 6.73 1.49 9.47 11.2 11.7 8.57 10.4 4.61 11.1 8.05 12.5 11.7 12.7 13.8 7.25 10.5 12.8 11.6 11.1 2.61 7.52 3.54 3.31 6.73

6.8 21.8 11.7 7.72 14.1 0.90 3.13 21.3 15.7 5.56 17.8 1.10 2.86 5.25 12.3 13.5 17.5 9.12 2.93 0.93 4.05 6.41 11.5 8.84 17.0 7.88 8.04 6.35 21.5 14.2 9.80 16.7 7.97 4.15 2.20 0.34

6.5 3.19 6.55 6.42 5.00 6.19 14.2 17.5 5.52 2.62 11.1 0.29 3.91 2.30 7.44 7.94 6.40 7.86 7.20 3.41 13.4 4.66 8.60 2.24 7.67 12.4 6.68 10.5 4.65 7.65 11.8 10.6 0.39 2.91 6.13 7.39

Source : Authors’ estimates.

individual LDCs’ and small states’ merchandize and commercial services exports over different periods. It shows only ﬁve LDCs (viz. Benin, Cambodia, Lao PDR, Uganda and Yemen) and eight small states (viz. Botswana, Equatorial Guinea, Lesotho, Maldives, Malta, Mauritius, Seychelles, and St Vincent and Grenadines) with growth rates above the global average.23

23 Of these countries some have been resource rich such as Botswana or have commenced from a very low base. Only in the case of Mauritius has there developed a diversiﬁed and merchandise export-led growth, which has provided a genuine counter-example to the inexorable tendency of small state marginalization.

195

The Implications of Declining Commodity Prices In contrast to merchandise trade the commercial services’ export performance of small states differs from that of LDCs. Only six LDCs enjoyed growth rates higher than that of the 1980–2000 trend world growth rate, while as many as 21 small states register a growth rate higher than the world average.24 Considering both the merchandise and services exports together, only 9 LDCs are found to have prevented their marginalization over the long period of 1980–2000 by registering positive growth in their share (see Figure 7.10). Cambodia turns out to be the best performing LDC achieving growth rates higher than the world average for both the merchandise goods as well as services exports. Three countries with civil strife viz. Afghanistan, Sierra Leone, and Rwanda appear to have become most marginalized during the past two decades. Considering the merchandise exports alone, only eight small states register positive trend growth in their share during the period of 1970–2000. But due to a better performance in the exports of commercial services the number of small states that have prevented marginalization during 1980–2000 rises to 14 (see Figure 7.11). According to Figure 7.11 Equatorial Guinea is the best performer while Trinidad and Tobago appears to have been the most marginalized small state. The relatively good performance of Equatorial Guinea stems from its very low base of merchandise exports in 1980 (see Appendix Figure 7.4).

7.3.2. Recent Performance of Individual Countries Appendix Tables 7.15–7.17 evaluate country speciﬁc performances in the 1990s. First, information on average change in merchandize exports for the early- (1990–94) and late-1990s (1995–2000) across countries is given in Appendix 7.15. Based on average absolute change for 1995–2000 the top ten countries emerged to be the United States, China, Mexico, Canada, Republic of Korea, the United Kingdom, Russian Federation, France, Taiwan and the Netherlands in that order. Together these ten countries account for about 53 per cent of annual average change in world merchandise exports. Adding the next ten top performers (viz. Japan, Ireland, Hong Kong, Germany, Saudi Arabia, Malaysia, Spain, Philippines, United Arab Emirates, and Singapore) to the above list help to explain 73 per cent change in world exports. Six HPAEs (viz. China, South Korea, Malaysia, Taiwan, Indonesia, and Thailand) and Mexico are responsible for two-thirds of the total change in merchandise goods exported by the developing world. On the other hand, LDCs were responsible for a mere 1.94 per cent of the change in the absolute volume for the developing world and SVs only 0.7 per cent.

24 In the following decades of the 1980s and 1990s, these numbers fall slightly to 19 and 17 respectively.

196

Percentage change in share 8

6

4

2

0

−4

Sao Tome and Principe

−15

−20 Afghanistan

Mali

Liberia Somalia Zambia

Mauritania Ethiopia Madagascar

Togo

Chad Tanzania Angola Mozambique Guinea Uganda Congo, Dem. Rep. of Central African Rep. Malawi Senegal

Sudan Niger Burundi Haiti Sierra Leone Rwanda

−10

Gabon Bahrain Suriname Trinidad and Tobago

−6

Papua New Guinea Belize Solomon Islands Jamaica Kiribati Guyana

−5

Fiji Vanuatu Barbados

−2

Swaziland Comoros Samoa St. Vincent & the Grenadines

Cape Verde St. Lucia Grenada St.Kitts and Nevis

0

Gambia Tonga Djibouti

10

Botswana Seychelles

12

Lesotho Cyprus

14 Benin Bangladesh Nepal Myanmar Yemen, Republic of Guinea-Bissau Burkina Faso

5

Antigua and Barbuda Dominica Malta

10 Cambodia Bhutan Lao PDR

15

Equatorial Guinea

20

Maldives Mauritius

Percentage change in Share

Marginalization and World Trade

Figure 7.10. Marginalization of Individual LDCs in Aggregate Exports (Merchandise plus Commercial Services), 1980–2000

Note: Based on authors’ estimates.

−8

Figure 7.11. Marginalization of Individual Small States in Total Exports (Merchandise plus Commercial Services), 1980–2000

Note: Based on authors’ own estimates.

197

The Implications of Declining Commodity Prices Appendix 7.16 provides similar information on commercial services exports. The United States, the United Kingdom, Spain, Ireland, China, India, Canada, Greece, Belgium-Luxembourg, and Hong Kong, are the top ten performing countries accounting for approximately 74 per cent of the change in the volume of services exports in the late 1990s.25 LDCs and small states have an annual average change of $157 million and $490 million respectively, which are only about 1.1 and 3.3 per cent of total change accrued to the developing world. Finally, it can be calculated from the information given in Appendix 7.17 that the annual average change in total exports of LDCs and small states in the late 1990s has been $2553 million and $1361 million respectively (i.e. 1.85 and 0.9 per cent of the change in developing countries’ total exports volume). Therefore, the growth in world export trade had largely concentrated around a few developed and advanced developing countries and had not occurred in such a way that would have helped constrain the widening gap between the relatively advanced and poorest countries.26 Table 7.13 and 7.14 provide individual LDCs’ and small states’ annual average change in merchandise and commercial exports during the early1990s (i.e. 1990–94) and late-1990s (i.e. 1995–2000). Among the LDCs, Angola, Bangladesh, Congo, D.R., Myanmar, Nepal, Sudan, and Yemen have the biggest positive changes in merchandise exports in the late-1990s.27 On the other hand, there are 14 countries that have experienced a negative average change in exports, or in other words their 1995–2000 average merchandise exports have fallen absolutely. LDCs performed very badly in terms of changes in services exports. Not only is the base of services exports for most countries low, the gains had been small as well. In the case of small states, with the exception of the oil-rich countries, Equatorial Guinea and Malta had posted some considerable positive average changes in the late-1990s. As many as 12 countries, as listed in Table 7.14, conceded a fall in average exports during the same time. Many small states, however, did relatively well in terms of average change of commercial exports both in the early- and late-1990s.28 Average absolute change in exports, however, may not be useful in evaluating a country’s relative importance in world exports. Even if there is an absolute increase in export volume, a country can still be deﬁned as being 25 Consideration of the top twenty countries would have accounted for 94 per cent of the change in services exports. 26 The low-income countries that beneﬁted most from the growth in world trade are China, India and Vietnam as these countries had recorded large positive average changes in their exports. However, similar experiences for most LDCs and small states remained unrealized. 27 Of these Angola, Sudan, and Yemen are actually oil-exporting countries, as mentioned earlier, and Bangladesh, Nepal, and Myanmar have transformed them as manufacturing exporters. 28 In fact, there are only three countries, viz. Papua New Guinea, Suriname, and Swaziland that have experienced negative average changes in the exports of commercial services in the late-1990s.

198

Marginalization and World Trade Table 7.13. Average Change in Exports of LDCs in the 1990s Merchandise Exports ($ Million)

Countries

Services Exports ($ Million)

90–94 95–2000 Annual Annual 90–94 95–2000 Annual Annual avg. avg. avg. change avg. change avg. avg. avg. change avg. change exports exports 90–94 95–2000 exports exports 90–94 95–2000

Afghanistan 328.4 140.8 Angola 3381.8 3888.2 Bangladesh 2007.9 4646.2 Benin 347.0 430.7 Bhutan 70.6 113.0 Burkina Faso 99.5 257.3 Burundi 84.3 66.8 Cambodia 267.6 740.8 Central Af. Rep. 107.0 164.9 Chad 168.7 227.3 Congo, DR 1044.0 1702.8 Ethiopia 245.3 492.1 Guinea 629.0 845.8 Guinea-Bissau 18.8 40.2 Haiti 112.5 142.4 Lao PDR 169.9 331.4 Liberia 334.0 480.3 Madagascar 306.3 270.7 Malawi 388.9 470.5 Mali 365.3 512.2 Mauritania 384.2 460.2 Mozambique 143.2 222.4 Myanmar 532.0 1007.4 Nepal 314.9 502.8 Niger 287.0 304.2 Rwanda 70.7 62.1 Senegal 726.8 960.5 Sierra Leone 133.1 22.0 Somalia 127.0 130.8 Sudan 387.6 716.7 Tanzania 411.8 669.0 Togo 252.1 418.5 Uganda 216.5 523.7 Yemen Rep. 703.0 2543.6 Zambia 980.1 902.5

20.0 223.0 202.9 27.5 2.3 11.2 11.5 101.0 7.7 10.0 5.5 18.7 8.5 3.5 19.5 55.5 5.0 12.4 18.7 6.0 22.0 7.7 118.4 39.6 14.3 22.9 7.3 5.5 3.5 37.3 47.1 15.1 64.1 60.5 95.4

6.0 600.8 598.4 8.8 11.6 9.7 11.1 17.0 0.8 12.1 343.0 17.1 45.0 6.2 10.7 0.8 20.0 23.9 7.9 13.6 17.4 13.5 108.0 91.8 6.4 0.2 13.2 5.8 7.0 120.0 3.7 9.8 11.9 451.0 56.2

4.8 103.2 377.4 117.4 n.a. 38.4 7.0 48.0 16.6 18.8 145.2 247.0 6.2 6.4 n.a. 35.4 41.0 146.6 31.4 60.8 15.0 157.2 147.8 277.6 17.4 24.4 313.0 52.6 n.a. 77.8 241.2 80.2 42.8 106.0 75.0

7.2 151.0 293.2 136.8 n.a. 43.3 3.0 130.3 11.5 29.7 132.8 345.5 6.7 4.7 n.a. 92.2 46.0 263.8 35.3 65.3 24.7 280.0 481.0 560.5 12.3 27.7 345.7 76.3 n.a. 44.8 568.5 69.2 159.7 148.5 77.3

1.0 21.2 30.7 1.2 n.a. 1.0 0.2 0.7 0.0 0.0 42.2 1.2 0.5 0.5 n.a. 11.2 2.5 13.5 3.7 5.0 0.7 22.0 40.7 90.2 3.5 2.2 11.7 10.2 n.a. 22.5 70.0 15.5 16.0 5.5 4.5

0.2 7.4 37.2 0.8 n.a. 0.8 0.4 11.2 0.6 1.6 1.0 15.4 1.2 0.8 n.a. 8.6 0.8 19.0 2.4 0.0 2.0 16.6 31.8 36.4 0.2 5.6 3.0 3.0 n.a. 11.6 9.8 2.4 16.4 4.0 0.6

Note : Small LDCs are not included in the above. Source : Authors’ estimates based on UNCTAD and WTO data.

marginalized. To avoid marginalization, exports must grow at a rate at least as high as that of world export trade. Appendix Tables 7.18 and 7.19 estimate average merchandise and commercial exports of individual countries for 1985–89, 1990–94 and 1995–2000 and their resultant shares, based on which Tables 7.15 and 7.16 classify the LDCs and small states according to whether their individual shares in both the early and late 1990s have increased, decreased, or have experienced a mixture of both. Focusing on merchandise exports of LDCs, as given in Table 7.15, it is found that while 17 countries (about 46 per cent) out of a total of 37 experienced falling shares in both

199

The Implications of Declining Commodity Prices Table 7.14. Average Change in Exports of Small States in the 1990s Merchandise Exports ($ Million)

Countries Antigua and Barbuda Bahrain Barbados Belize Botswana Cape Verde Comoros Cyprus Djibouti Dominica Equatorial Guinea Fiji Gabon Gambia Grenada Guyana Jamaica Kiribati Lesotho Maldives Malta Mauritius Papua New Guinea Samoa Sao Tome and Principe Seychelles Solomon Islands St Kitts and Nevis St Lucia St Vincent and Grens. Suriname Swaziland Tonga Trinidad and Tobago Tuvalu Vanuatu

Services Exports ($ Million)

Annual 90–94 95–2000 Annual Annual 90–94 95–2000 Annual avg. avg. change avg. change avg. change avg. change avg. avg. avg. exports exports 90–94 95–2000 95–2000 exports exports 90–94 48.3

40.2

5.9

2.9

342.6

395.5

20.8

11.6

3616.1 4384.9 196.2 265.0 113.8 161.8 1795.9 2393.7 5.1 12.0 19.6 11.1 945.0 1147.3 19.0 20.7 51.7 53.3 69.8 354.6

35.9 8.39 4.7 15.8 0.2 1.7 4.5 1.2 1.9 0.4

317.5 6.7 10.3 21.6 0.4 0.1 55.4 0.4 1.5 59.5

556.8 715.0 652.0 954.2 101.0 130.2 177.2 251.3 32.4 78.3 15.2 36.5 1997.6 2725.5 33.6 27.2 41.6 78.8 6.2 6.7

115.0 39.5 6.5 2.0 2.2 3.7 159.5 0.5 4.5 0.5

29.4 42.2 7.0 23.4 8.4 3.8 43.0 0.0 5.0 1.2

482.6 2234.7 45.6 22.8 333.1 1118.8 4.8 102.8 45.3 1370.1 1267.2 1956.2

613.9 3127.9 15.9 23.2 532.2 1340.4 6.5 181.5 66.0 1904.2 1618.5 2189.8

18.8 36.6 1.1 1.0 49.8 13.5 0.5 20.4 1.5 110.0 38.2 363.6

4.5 233.9 1.47 0.25 23.0 26.2 0.2 3.9 5.3 84.5 8.4 111.5

419.8 540.0 260.6 210.8 65.4 91.8 79.2 118.8 100.8 143.8 1129.8 1749.0 13.0 19.2 32.0 43.7 142.8 306.5 846.2 1101.3 552.6 937.5 267.2 332.7

28.0 3.5 5.7 9.2 8.0 119.0 2.2 1.0 23.5 63.5 37.8 9.2

6.6 6.0 15.6 9.4 7.2 84.0 0.4 1.2 23.0 12.2 58.8 8.2

6.2 14.4

13.9 8.2

1.3 5.0

51.4 105.6

125.6 145.0

1.3 17.9

26.0

25.5

1.5

117.3 67.2

75.6 46.8

572.2 488.8 649.6 932.7 13.3 12.0 1833.0 2767.2 n.a. 21.7

n.a. 29.9

1.1 0.2

33.4 4.4

56.7 8.7

1.5 0.5

1.6 1.2

177.4 30.0

253.7 51.0

6.0 5.7

21.0 4.4

2.2

75.2

92.2

9.5

2.6

5.2 8.1

12.8 0.9

190.2 52.6

292.8 103.3

22.0 5.0

8.6 10.4

5.9 55.9 0.6 23.4

8.4 14.9 0.6 317.7

43.0 95.6 12.6 364.6

84.5 106.5 14.8 513.2

9.7 2.0 0.7 1.2

3.2 15.6 0.4 59.4

n.a. 1.6

n.a. 0.5

n.a. 63.0

n.a. 95.2

n.a. 3.5

25.4 15.07

Source : Authors’ estimates based on the data from the UNCTAD and the WTO.

200

n.a. 8.40

Marginalization and World Trade Table 7.15. A Summary of Trends in Marginalization of LDCs in the 1990s Avg. share increased both in 1990–94 (compared to 1985–90) and also in 1995–2000 (compared to 1990–94)

Avg. share fell both in Avg. share decreased 1990–94 Avg. share increased in 1990–94 (compared to in 1990–94 (compared to Avg. share decreased 1985–90) and (compared to 1985–90) but in 1995–2000 in 1995–2000 1985–90) but fell in increased in (compared to (compared to 1995–2000 1990–94) 1995–2000 1990–94) Merchandise Exports

Bangladesh Burkina Faso Ethiopia Lao PDR Mali Mozambique Myanmar Nepal

Angola Benin* Chad*

Bhutan* Cambodia* Guinea-Bissau Sudan* Tanzania* Togo* Uganda*

Afghanistan Angola Benin Burundi Central Af. Rep. Chad Congo Guinea Haiti Liberia Madagascar Malawi Mauritania Niger Rwanda Senegal Sierra Leone Somalia Yemen Zambia

Afghanistan Burundi Central Af. Rep. Congo Guinea Haiti Liberia Madagascar Malawi Mauritania Niger Rwanda Senegal Sierra Leone Somalia Yemen Zambia

Afghanistan Angola Bangladesh Benin Burkina Faso Burundi Central Af. Rep. Congo Ethiopia Guinea Liberia Malawi Mali Niger Rwanda Senegal Sudan Togo Yemen Zambia

Afghanistan Angola Bangladesh Central Af. Rep. Congo Ethiopia Guinea Liberia Malawi Mali Niger Rwanda Senegal Sudan Togo Yemen Zambia

Commercial Services Cambodia Lao PDR Madagascar Mozambique Myanmar Nepal Tanzania Uganda

Benin Burkina Faso* Burundi* Guinea-Bissau Sierra Leone

Chad* Haiti Mauritania*

(Continued )

201

The Implications of Declining Commodity Prices Table 7.15. (Continued ) Avg. share increased both in 1990–94 (compared to 1985–90) and also in 1995–2000 (compared to 1990–94)

Avg. share fell both in Avg. share decreased 1990–94 Avg. share increased in 1990–94 (compared to in 1990–94 (compared to Avg. share decreased 1985–90) and (compared to 1985–90) but in 1995–2000 in 1995–2000 1985–90) but fell in increased in (compared to (compared to 1995–2000 1995–2000 1990–94) 1990–94) Total exports (Merchandise Plus Services)

Bangladesh Cambodia Lao PDR Myanmar Nepal

Benin Mali*

Burkina Faso Congo* Ethiopia* Guinea-Bissau Mozambique Sudan Tanzania Yemen

Afghanistan Angola Benin Burundi Central Af. Rep. Chad Guinea Haiti Liberia Madagascar Malawi Mali Mauritania Niger Rwanda Senegal Sierra Leone Somalia Togo Zambia

Afghanistan Angola Burundi Central Af. Rep. Chad Guinea Haiti Liberia Madagascar Malawi Mauritania Niger Rwanda Senegal Sierra Leone Somalia Togo Zambia

Note: * indicates that average share was lower than that of 1985–90. For Bhutan information on services exports is not available and hence it is not shown in the rows for commercial and total exports. No information on Tuvalu and Eritrea is available. 12 LDCs that are also small states are not considered here but are classiﬁed in the group of small states. Therefore, there is a total of 34 LDCs in the above table. Source: Based on Authors’ own computation.

halves of the 1990s, there are even more countries (20) with a declining share in the late-1990s.29An almost similar picture is obtained for commercial services. Considering both the merchandise goods and commercial services together while there are only ﬁve LDCs, viz. Bangladesh, Cambodia, Lao PDR, Myanmar, and Nepal, that have been able to increase their shares both in the early and late 1990s, as many as 20 countries conceded declines consecutively. For small states, as Table 7.16 shows, almost two-thirds (22 out of a total of 35) have seen their average share in merchandise exports fall in 1990–94 and 1995–2000. There are 27 countries (77 per cent) with declining shares in the post-Uruguay Round of 1995–2000. The only two countries that could

29 Note that 12 LDCs that are also small states are not considered here. Small-LDCs are included in Table 7.16.

202

Marginalization and World Trade Table 7.16. A Summary of Trends in Marginalization of Small States in the 1990s Avg. share increased both in 1990–94 (compared to 1985–90) and also in 1995–2000 (compared to 1990–94)

Avg. share fell both in Avg. share decreased 1990–94 Avg. share increased in 1990–94 (compared to in 1990–94 (compared to Avg. share decreased 1985–90) and (compared to 1985–90) but in 1995–2000 in 1995–2000 1985–90) but fell in increased in (compared to (compared to 1995–2000 1990–94) 1995–2000 1990–94) Merchandise Exports

Equatorial Guinea Seychelles

Malta Mauritius* Papua New Guinea* Swaziland Tonga*

Antigua and Barbuda* Cape Verde* Guyana Samoa* Trinidad and Tobago*

Bahrain Barbados Belize Botswana Comoros Cyprus Djibouti Dominica Fiji Gabon Gambia Grenada Jamaica Kiribati Maldives Malta Mauritius Papua New Guinea Sao Tome and Principe Solomon Islands St Kitts and Nevis St Lucia St Vincent and Grenadines Suriname Swaziland Tonga Vanuatu

Bahrain Barbados Belize Botswana Comoros Cyprus Djibouti Dominica Fiji Gabon Gambia Grenada Jamaica Kiribati Maldives Sao Tome and Principe Solomon Islands St Kitts and Nevis St Lucia St Vincent and Grenadines Suriname Vanuatu

Antigua and Barbuda Bahrain Barbados Belize Botswana Cyprus Djibouti Equatorial Guinea Fiji

Antigua and Barbuda Bahrain Barbados Djibouti Equatorial Guinea Gambia Guyana Seychelles Tonga

Commercial Services Comoros Dominica Grenada Kiribati Maldives Mauritius Namibia Samoa Sao Tome and Principe Solomon Islands

Belize Botswana Cyprus Fiji* Gabon* Lesotho* Malta* Papua New Guinea St Kitts and Nevis*

Cape Verde Jamaica* Suriname*

(Continued )

203

The Implications of Declining Commodity Prices Table 7.16. (Continued ) Avg. share increased both in 1990–94 (compared to 1985–90) and also in 1995–2000 (compared to 1990–94) St Lucia St Vincent and Grenadines Vanuatu

Avg. share fell both in Avg. share decreased 1990–94 Avg. share increased in 1990–94 (compared to in 1990–94 (compared to Avg. share decreased 1985–90) and (compared to 1985–90) but in 1995–2000 in 1995–2000 1985–90) but fell in increased in (compared to (compared to 1995–2000 1995–2000 1990–94) 1990–94) Swaziland*

Gabon Trinidad and Gambia Tobago Guyana Lesotho Malta Papua New Guinea Seychelles St Kitts and Nevis Swaziland Tonga Trinidad and Tobago

Total exports (Merchandise Plus Services) Equatorial Guinea Lesotho Maldives

Antigua and Barbuda Belize* Cyprus Malta Mauritius Papua New Guinea* Solomon Islands St Lucia* Swaziland

Cape Verde Guyana Samoa* Seychelles

Note: * indicates that average share was lower than that of 1985–90. Source: Based on authors’ own computation.

204

Antigua and Barbuda Bahrain Barbados Belize Botswana Comoros Cyprus Djibouti Dominica Fiji Gabon Gambia Grenada Jamaica Kiribati Malta Mauritius Papua New Guinea Sao Tome and Principe Solomon Islands St Kitts and Nevis St Lucia St Vincent and Grenadines Suriname Swaziland Tonga Trinidad and Tobago Vanuatu

Bahrain Barbados Botswana Comoros Djibouti Dominica Fiji Gabon Gambia Grenada Jamaica Kiribati Sao Tome and Principe St Kitts and Nevis St Vincent and Grenadines Suriname Tonga Trinidad and Tobago Vanuatu

Marginalization and World Trade prevent marginalization throughout the 1990s were Equatorial Guinea and Seychelles.30 In the case of commercial services, small states’ performance is much better; there are as many as thirteen countries (about 37 per cent) that have increased their shares both in 1990–94 and 1995–2000. Also, only eleven countries have become marginalized in both halves of the 1990s. However, the number of countries (21) that have been subject to declining share in the post-Uruguay Round is still comparable to that of merchandise exports. Finally, when merchandise and services are considered together, there remain only 3 countries, viz. Equatorial Guinea, Lesotho, and Maldives that survived marginalization in the 1990s. There are 28 countries (80 per cent) experiencing reduced share in the late 1990s of which 19 conceded decline in both periods.

7.3.3. Lost exports due to marginalization How much of the lost exports are as a result of marginalization? This may be estimated as the net shifts in exports resulting from the discrepancies between the actual export receipts and the predicted earnings based on countries’ share at a previous point in time. A positive net shift is associated with a country’s rising share while a negative shift reﬂects diminishing relative importance or marginalization. It thus shows, for example, if Antigua and Barbuda could maintain its average 1990–94 share in world exports it should have had average export receipts of $578 millions in 1995–2000. In reality, its average earnings in the late-1990s stood at $436 millions thus registering a net negative shift of -$142 millions, or about 36 per cent of its average 1990–94 export earnings. Figures 7.12 and 7.13 exhibit the net gains attributable to total exports of LDCs and small states in 1995–2000 as a percentage of their 1990–94 exports. In the case of LDCs, twenty countries suffered negative shifts, while seventeen experienced positive shifts. As quite a few LDCs gained between 1995 and 2000, LDCs managed to escape further deterioration in their relative importance, as shown in Figure 7.1, during the second half of the 1990s. Gains to Cambodia, Uganda and Yemen have been quite impressive.31 Besides, since in 1990–94 LDCs’ share was already very low, net negative shifts in the late 1990s appeared to be not very large. However, if the individual countries’ share for the early 1980s (1980–85) were considered, LDCs as a group should have an average export of about $46 billion during 1995–2000 as against its actual exports of about $32 billion in 2000.32 30 The growth of exports of cocoa beans in the case of Equatorial Guinea and canned tuna for Seychelles resulted in such favourable outcomes. 31 For Yemen the gains are due to a rise in oil exports. Reinvigoration of trading activities after a long period of economic stagnation is the reason for the success of Uganda and Cambodia. 32 Appendix 20 gives individual LDCs’ average net shifts in 1995–2000 in comparison with their predicted exports on the basis of 1980–5 average share.

205

Per cent

200

150

100

50

206 250

350

300

Equatorial Guinea

Bangladesh Lao PDR Burkina Faso Myanmar Cambodia

Nepal Tanzania

Ethiopia Guinea-Bissau

Sudan Mozambique

200

Samoa Maldives Cape Verde

Vanuatu Trinidad and Tobago Guyana Seychelles Lesotho

Barbados Kiribati Solomon Islands

Congo, Dem. Rep. of Bhutan

Central African Rep. Togo

Mali Chad Liberia

Zambia Guinea

Haiti Senegal

150

Mauritius Dominica

Malta Belize Comoros Jamaica Grenada Swaziland

St Vincent and the Grenadines Fiji Cyprus Gabon Botswana

Rwanda Niger Somalia Angola Madagascar Malawi Mauritania Benin

0

Gambia Tonga Antigua and Barbuda Papua New Guinea St Kitts and Nevis St Lucia Bahrain

50 Afghanistan Sierra Leone Burundi

100

Sao Tome and Principe Djibouti Suriname

Per cent

350

300 Uganda Yemen

The Implications of Declining Commodity Prices

Turning to the small states, it is observed that Sao Tome and Principe, Djibouti, Suriname, and Gabon experienced negative net shifts in excess of 50 per cent of their average 1990–94 exports. Tonga, Antigua and Barbuda, Papua New Guinea, St Kitts and Nevis, St Lucia, Bahrain, St Vincent and the

250

Figure 7.12. Net Shifts in 1995–2000 as Percentage of 1990–4 Average Exports (Merchandise Plus Services) for LDCs

0

Figure 7.13. Net Shifts in 1995–2000 as Percentage of 1990–94 Average Exports (Merchandise Plus Services) for Small States

Marginalization and World Trade Grenadines, and Fiji, are countries that have negative shifts between 20 and 50 per cent of their average exports of 1990–94. Only seven small states are found to have enjoyed net positive shifts, of which gains to Cape Verde and Equatorial Guinea are remarkably high. For the whole group of small states net shifts were negative and were estimated to be 15 per cent of their average 1990–94 export earnings. Therefore, in order to prevent marginalization just in the post-Uruguay Round period small

14

y = 0.4007x + 4.2585 R2=0.5052

GDP growth rate (per cent)

12

Equatorial Guinea Maldives

10 8 Cape Verde Solomon Islands

6 Gambia

4

Bhutan

Lesotho

2 0 Congo. D.R. −2

Angola Kiribati

−4 −15

−10

−5 0 Trend growth in export share (per cent)

5

10

Figure 7.14. Trends in Marginalization and Growth of Real GDP in LDCs

14

GDP Growth Rate = 0.6579 Growth of Exports Share + 4.1811 R2 = 0.6571

GDP growth rate

12

Equatorial Guinea

10 Maldives

Botswana

8 6 4 2

Bahrain Gambia Comoros

Suriname

Trinidad and Tobago

0

−2 −8

Dominica

Samoa SaoTome&Principe Djibouti

−6

−4

−2

0

2

4

6

8

10

12

14

Trend growth rates in export share

Figure 7.15. Trends in Marginalization and Growth of Real GDP in Small States Note: Growth rates of exports share are authors’ estimates. Except Equatorial Guinea, average GDP growth rates for 1980-99 are taken from UNCTAD (2002). For Equatorial Guinea the average growth rate of GDP has been taken from Commonwealth Secretariat, Small States: Economic Review & Basic Statistics (various issues).

207

The Implications of Declining Commodity Prices states would have required an additional $4.2 billion worth of exports of merchandise goods and commercial services per annum.33 Figures 7.14 and 7.15 examine how GDP growth rates are related to growth rates of share of exports over the period 1980–98 for LDCs and SVs respectively. Negative trend growth rates of exports show the marginalization of a country, while positive rates indicate increasing relative importance. A clear positive relationship is obtained in both ﬁgures suggesting that countries that have been able to improve their relative importance also enjoy higher GDP growth rates.

7.4. Marginalization in Merchandise Export Trade: A Statistical Analysis 7.4.1. Understanding marginalization The manufacturing export base in most LDCs and small states is rudimentary in nature and most of these countries have to rely overwhelmingly on agricultural commodities and natural resource intensive products for export. As can be seen from Figure 7.16 in 25 LDCs, out of a total of 33 for which the information is available, primary exports contribute to more than 50 per cent of the receipts from merchandise exports. Only in the cases of Madagascar, Sierra Leone, Haiti, Nepal, and Bangladesh do manufacturing exports dominate the export basket.34 Turning to small states, Figure 7.17 shows that for 24 small states (out of a total of 28 for which information is available), primary exports accounted for more than half of the total merchandise export earnings. The excessive reliance on primary commodities has grave implications for the long-term relative signiﬁcance of the countries that depend on them. As the income elasticity of demand for agricultural products is low, production of and trade in primary commodities have failed to keep pace with the growth of world trade. In 1980 agricultural products constituted about 16 per cent of world merchandise exports, whereas it now accounts for just seven per cent. This follows the classic Engel’s law, which explains the tendency of consumers to spend less on basic food products (or primary commodities) as their incomes rise. This process has been exacerbated by the development of new technology and improved productivity growth. The advent of new technologies reduces the intensity of the use of various primary commodities such as metals and 33 As explained in the case of LDCs, the estimates of net gains are sensitive to the choice of the time period for which average shares are considered to predict the export receipts. Appendix 21 provides net shifts in exports of individual small states in 1995–2000 on the basis of their average share during 1980–85. 34 With the growth in the exports of textiles and clothing products, Bangladesh and Nepal have witnessed a structural transformation whereby the dominance of primary commodities has been replaced by manufactured goods. While Sierra Leone is shown to have a little dependence on primary commodities, its manufacturing activities (e.g. diamond-cuts) are heavily dependent on natural resources.

208

Kiribati

Grenada

Malta

Mauritius

Cyprus Barbados

Dominica

Trinidad and Tobago

St Kitts and Nevis

Fiji

St Lucia

Jamaica

Maldives

Primary Exports

Comoros Gambia

Madagascar Cambodia Sierra Leone Haiti Bangladesh Nepal

Guinea Myanmar Lao PDR Togo Senegal

Niger Rwanda Maldives

Liberia Zambia Angola Mauritania Yemen, Republic of Guinea-Bissau Samoa Congo, D.R. Chad Benin Mali Uganda Mozambique Central Af. Rep. Malawi Sudan Burundi Burkina Faso Tanzania Primary Exports

Belize St Vincent & Grenadines

Vanuatu Bahrain

Togo

Gabon

Djibouti

Samoa Papua New Guinea

Seychelles Suriname

Tuvalu

Solomon Islands

Marginalization and World Trade

100 Manufacturing exports

90

80

70

60

50

40

30

20

10

0

Figure 7.16. Composition of Exports in LDCs: Primary vs. Manufacturing

100 Manufacturing exports

90

80

70

60

50

40

30

20

10

0

Figure 7.17. Share of Primary and Manufacturing Exports in Small States

Note: Data have been compiled from UNCTAD (2002b), World Bank (2000), and Commonwealth Secretariat (various issues). All data correspond to a year in the late 1990s.

209

The Implications of Declining Commodity Prices agricultural raw materials (World Bank, 1994) and is also responsible for productivity improvement and increased production of many agricultural commodities (Reinhart and Wickham, 1994) thereby depressing the demand but inducing supply, which interact to exert downward pressures on primary commodity prices. Therefore, the operation of Engel’s law (resulting in diminishing share of agricultural goods in total global consumption expenditure) would guide the effects of technology-induced decline in demand and the increased supply capacity is translated into depressed prices for agricultural commodities. Empirical evidence especially since the late 1970s corroborates this hypothesis. Between 1970 and 1993, real commodity prices more than halved (World Bank, 1994).35 Then, these prices registered an annual average growth of 6 per cent during 1994–97. This was followed by consecutive declines of 13 and 14.2 per cent respectively in 1998 and 1999 (UNCTAD, 2000). Table 7.17 shows that real prices of 17 major commodities in 2000 were lower than their corresponding prices in 1980 by 25 per cent or more. For another 8 commodities prices have fallen by more than 50 per cent in the past two decades. The falling prices of agricultural products have also certainly contributed to the marginalization of small states.36 It needs to be mentioned here that advances in technology and productivity are also taking place in the manufacturing sector. But manufacturing activities provide much more scope for product diversiﬁcation and innovation allowing individual countries to specialize in different market niches. Moreover, the income elasticity of demand for manufacturing exports, in general, is higher than primary commodities, i.e. a given rise in income will result in a proportionately higher expenditure on the former. All this implies that prices Table 7.17. Fall in Commodity Prices in Real Terms

Commodity

Fall in real price in 2000 in comparison to 1980 (%)

Banana Fertilizer Iron Ore Tea Aluminium Coconut Oil Copper

4.4 23.1 19.5 7.5 27.2 44.3 30.9

Commodity Cotton Fish meal Groundnut Oil Maize Soybean Wheat Cocoa

Fall in real price in 2000 in comparison to 1980 (%) 47.6 31.9 30.9 41.6 39.0 45.2 71.2

Commodity Coffee Lead Palm Oil Rice Rubber Sugar Tin

Fall in real price in 2000 in comparison to 1980 (%) 64.5 58.3 55.8 60.9 59.6 76.6 73.0

Source : Oxfam estimates from IMF International Financial Statistics Yearbook.

35 According to World Bank (1994: 32) ‘the estimated annual loss to developing countries from the fall in commodity prices between 1980 and 1993 reached $100 billion a year in 1993—or more than twice the total ﬂow of aid in 1990.’ 36 In fact, the prices of primary commodities have also fallen relative to manufacturing goods causing terms of trade shocks for countries specializing in agricultural products.

210

Marginalization and World Trade of primary commodities relative to those of manufactured goods will have a long-run tendency to deteriorate in the world markets.37 The other obvious driving force behind the diminishing share of LDCs and small states is the rapid rise in trading activities in the world economy itself in the present era of globalization. In particular, among all other categories, trade in high and medium technology-intensive products have registered the most dynamic growth trends in recent times capturing the largest share of world trade (Commonwealth Secretariat and UNCTAD, 2001), export of such products is virtually non-existent from LDCs and small states.38

7.4.2. A simple model of marginalization of small states The above discussions lead us to perceive the problem of marginalization of LDCs and small states in merchandise trade from two main perspectives. First, the overriding problem has been the overwhelming dependence on basic primary commodities. As the demand for these goods is income inelastic by nature, with a rise in world income there will be a natural tendency for primary exporters’ share to shrink. Other things remaining constant following Engel’s law primary producing countries will show a natural tendency of becoming marginalized in an expanding global economy. On the other hand, globalization has resulted in a rapid rise in world trade mostly dominated by manufactured goods. But due to their inability to transform their export base, LDCs and small states have not been part of this growth process. As globalization accelerates and the same raw materials cross more borders until they are ﬁnally processed into manufactured goods, countries conﬁned to the production of raw materials will experience marginalization. Thus, a simple model of marginalization of LDCs and small states can be written in the following way: MAR ¼ f (AGX, GLO)

(1)

where, MAR, AGX, and GLO stand, respectively, for marginalization (as measured by share of merchandise exports of LDCs and small states in total global merchandise exports), share of agricultural goods in total global merchandise exports and a measure of globalization. For this study the world export-GDP ratio will be considered as a measure of globalization. According to our hypothesis one should obtain a positive relationship between MAR and AGX but an inverse association between MAR and GLO. Using log-linear transformation and adding an intercept (a) as well as a stochastic error term (e) the estimating form of (1) becomes: ln MAR ¼ a þ b1 ln AGX þ b2 lnGLO þ «

(2)

37

This is usually known as the Singer–Prebisch theory. According to Commonwealth Secretariat and UNCTAD (2001) the value of trade in ofﬁce products now exceeds the value of agricultural trade. 38

211

The Implications of Declining Commodity Prices

7.4.3. Data The shares of LDCs and small states in world merchandise exports have already been calculated in section 7.2. The shares can be computed for a period of 51 years: 1950–2000. To estimate AGX, the data on world exports of agricultural products have been gathered from the Food and Agricultural Organization (FAO) Yearbook, which are available for a maximum of thirty years thus reducing the sample to 30 observations (1970–99). Using these data and UNCTAD information on world merchandise exports, the ratio of agricultural goods to merchandise exports is calculated. Finally, GLO is to be proxied by the world export-GDP ratio, which is not available for the whole period of 1970–99 from any secondary sources.39 Therefore, the series of world GDP was constructed by using the IMF index of world GDP volume and using the World Bank estimate of world GDP in 1998.40 Figure 7.18 provides graphical plots of the share of agricultural commodities in world merchandise exports and world exports-GDP ratio. In 1970 world Ratio of Agricultural to Merchandise Exports in world: 1970–99 0.2

World Exports-GDP ratio 0.24 0.22

0.18

0.2

0.16

0.18

ratio

ratio

0.14

0.16

0.12

0.14

0.1

0.12

0.08

1997

1994

1991

1988

1985

1982

1979

1976

1973

1997

1994

1991

1988

1985

1982

1979

1976

1973

1970

1970

0.1

0.06

Figure 7.18. Share of Agriculture in World Exports and World Exports-GDP Ratio 39 The information on world exports is reported in different sources, e.g. IMF, UNCTAD, and World Bank. However, the time series on world GDP cannot be obtained from any published database. 40 According to the World Bank (2000) world GDP in 1998 stood at $28445 billion. Then using IMF index of world GDP volume, as given in the International Financial Statistics (IFS), the ﬁgures for other years were constructed.

212

Marginalization and World Trade exports of agricultural products stood at $53.5 billion, which by the end of the 1990s rose to $442 billion: an eight-fold increase over 1970. During the same time, however, the total world merchandise exports grew by more than eighteen times despite the fact that even in 1970 the base of overall merchandise exports was about six times higher than that of agricultural exports. As a result, there has been a secular decline in the share of agriculture in total exports from about 18 per cent in the beginning of the sample to just 8 per cent in the late 1990s, which is being illustrated vividly in the left panel of Figure 7.11. The world-export GDP ratio has a rising trend during the past three decades but the trend is characterized by wild ﬂuctuations as reﬂected in the right panel of Figure 7.11. In the 1970s the world exports-GDP ratio doubled from just over 11 per cent in 1970 to about 23 per cent in 1980 as over the decade merchandise exports registered a staggering trend growth rate of 19 per cent. Then in the early 1980s exports grew at much slower rates, even marked by negative growth rates (decline in absolute exports), which led to a fall in the export ratio. However, from the mid-1980s the trend in world exports-GDP ratio was reversed upward again with some notable ﬂuctuations.

7.4.4. Empirical estimation of the model for LDCs First we will estimate the model for LDCs. Figures 7.19 and 7.20 provide the scatter plots of the dependent and independent variables in the model (2). As expected, a positive relationship between lnMAR and lnAGX is observed in Figure 7.19 and an inverse relationship between lnMAR and lnGLO in Figure 7.20. Although these two ﬁgures demonstrate the bivariate relationship between the dependent and explanatory variables, we need to estimate the model carefully to ascertain that there is a ‘genuine’ long-run relationship between them. −2.7

−2.5

−2.3

−2.1

−1.9

lnMAR = 1.5161 lnAGX - 1.8203 R 2 = 0.8577

−1.7

−1.5 −4 −4.2 −4.4 −4.6

lnMAR

−4.8 −5 −5.2 −5.4 −5.6

lnAGX

−5.8

Figure 7.19. Scatter Plot of lnMAR and lnAGX

213

The Implications of Declining Commodity Prices −2.3

−2.2

−2.1

−2

−1.9

−1.8

−1.7

−1.6

−1.5

lnMAR = −1.2094 lnGLO −7.1794 R 2 = 0.2302

−1.4 −4 −4.2 −4.4 −4.6

lnMAR

−4.8 −5 −5.2 −5.4 −5.6 lnGLO

−5.8

Figure 7.20. Scatter Plot of lnMAR and lnGLO

7.4.4.1. TESTS FOR UNIT ROOTS AND COINTEGRATION The model in equation (2) postulates a long-run relationship between the dependent and explanatory variables based on the time series data. The relatively recent developments in time series econometrics have demonstrated the problem of the Ordinary Least Squares (OLS) regression in estimating such models containing variables that are not stationary. A time series is said to be stationary if its mean, variance and auto-covariances are independent of time. By now there is compelling evidence that many time series are non-stationary and the use of OLS in estimating regression coefﬁcients may produce spurious results. Under such circumstances the validity of the long-run relationship between the variables in the model may be questioned.41 In other words, non-stationary time series can produce spurious correlation. In order to avoid this problem it is necessary to consider the integrating properties of the variables and to use an appropriate estimation strategy. Graphical plots of the variables as given in Figure 7.18 give a ﬁrst hand impression that the variables in our model may be non-stationary on their levels. However, formal tests should be employed in determining the integrating properties of the time series and to distinguish between non-stationary and stationary variables. These tests are known as the Unit root tests. 7.4.4.2. TESTING THE VARIABLES FOR UNIT ROOTS First, we perform an F-test of the following form on MAR, AGX, and GLO with the joint restrictions that c ¼ 1 and x ¼ 0: 41

An OLS regression involving non-stationary variables resulting in high R2 can also be misleading. Moreover, the estimated standard errors and test statistics for ‘t’ and F-tests become non-standard providing invalid inferences.

214

Marginalization and World Trade DYt ¼ t þ (c 1)Yt1 þ xT þ et

(3)

where Y is the variable under consideration, D is the ﬁrst difference operator, subscript t denotes time period, T is the time trend and e is the error term. The null hypothesis in this case is that the series is non-stationary (i.e. it contains a unit root) against the alternative of stationary. The F-test values obtained for these series are presented in Table 7.18. It can now be seen that the null hypothesis of non-stationary can be rejected for none of the variables as in every case the computed F-statistic falls short of its critical value. Although the above test indicates non-stationarity of the variables, the popular methods for testing unit roots are the Dickey-Fuller (DF) and the Augmented Dickey-Fuller (ADF) tests. The DF test for unit root is also based on equation (3) with the null hypothesis of (ł 1) ¼ 0 (i.e. Yt is non-stationary) against the alternative of (ł 1) < 0 (i.e. Yt is stationary). The Augmented DickeyFuller (ADF) test, on the other hand, is a modiﬁcation of the DF test and involves augmenting equation (3) by lagged values of the dependent variables. This is done to ensure that the error process in the estimating equation is residually uncorrelated. Thus, the ADF version of the test is based on the following equation: DYt ¼ t þ (c 1)Yt1 þ xT þ dDYt1 þ et

(4)

The t-test on the estimated coefﬁcient of Yt1 in equations (3) and (4) provides the DF and ADF test statistics for the presence of a unit root. In both cases, the estimated t-ratios are non-standard and thus the computed statistics need to be compared with the corresponding critical values to infer about the stationarity of the variables. It is quite common to ﬁnd that many time series are non-stationary on their levels but stationary on their ﬁrst or higher order differences. Following Engle and Granger (1987) a non-stationary series is said to be integrated of order d, usually denoted as ~I(d), if the series can be transformed into a stationary process by differencing it d times. Note that although in equations (3) and (4) the trend term T is included, most studies in applied time series econometrics report the DF-ADF test results by including and excluding the trend term separately. When a variable is Table 7.18. Computed F-Test Statistics and Critical Values Variables lnMAR lnAGX lnGLO

Computed F

Critical F

7.65 2.84 4.95

10.61 10.61 10.61

Remark the variable is non-stationary the variable is non-stationary the variable is non-stationary

Note : In this case the F-test is non-standard and the appropriate critical values are given by Dickey and Fuller (1981) as cited in Maddala (1992).

215

The Implications of Declining Commodity Prices Table 7.19. DF and ADF Tests for Unit Roots DF-ADF tests without the trend term

DF-ADF tests including the trend term

Variables

DF

ADF

DF

ADF

lnMAR D lnMAR lnAGX D lnAGX lnGLO D lnGLO

3.73* 4.58* 0.60 5.65* 3.34* 3.13*

3.29* 3.94* 0.54 4.35* 3.98* 3.14*

2.51 5.10* 3.08 5.54* 3.06 3.22

2.48 5.02* 3.19 4.27* 3.73* 3.29

Note: (1) The 95 per cent critical values for DF-ADF tests with and without the trend term are 2.97 and 3.57 respectively. (2) D implies ﬁrst difference of the respective variables. (3) * indicates rejection of the null-hypothesis of non-stationarity at the 95 per cent level of statistical signiﬁcance.

trended, the test statistics, including the trend term, is preferred. In the case of conﬂicting results, ADF test is preferred to DF. Table 7.19 gives the results of the DF and ADF tests. For lnMAR the computed DF and ADF test statistics without the trend term are absolutely higher than the 95 per cent critical value whereas the same tests with the trend term provide completely different result. Since lnMAR is strongly trended, it is more appropriate to consider the test results with the trend term, which suggests that lnMAR is non-stationary on its level. The unit root tests for ﬁrst difference of lnMAR (DlnMAR), however, overwhelmingly reject the null hypothesis of non-stationarity as all the test statistics exceed the 95 per cent critical value (absolutely). Thus we conclude that while lnMAR is non-stationary on its level but stationary on its ﬁrst differences, or lnMAR is integrated of order one: i.e., lnMAR ~I(1). The results for lnAGX are rather robust as all tests indicate its non-stationarity and stationarity of DlnAGX. That is, lnAGX ~I(1). Finally, the graph of lnGLO does not seem to be trended and thus considering DF and ADF tests without the trend term it can be regarded as a stationary variable on its level [lnGLO ~I(0)]; the ADF test with the trend term also supports such a conclusion. Thus the tests seem to suggest that lnMAR and lnAGX are ~I(1), while lnGLO is ~I(0). It needs to be mentioned that the low power of DF and ADF tests is well acknowledged in the literature and the most important problem faced when applying the tests is their probable poor size and power properties. This is often reﬂected in the tendency to over-reject the null hypothesis when it is true and under-reject when it is false (Harris, 1995). This problem is particularly severe in the case of a small sample like ours. Thus Hall (1986) emphasizes the importance of the inspection of the autocorrelation function and correlogram in determining the integrating properties of the variables with short time length. The correlograms of the level and ﬁrst differenced variables are given in Figure 7.21 side by side the graphical plots of the variables.42 For a 42 These correlograms are graphical plots of autocorrelation functions of individual variables. The autocorrelation function at lag k, denoted as pk, is deﬁned as the ratio of covariance

216

Marginalization and World Trade

−4

lnMAR

lnMAR

DlnMAR

1

−4.5

0 .5

−5

−.1

−5.5 1970 1

−.2 1980

1990

2000

5

DlnMAR

10

lnAGX

1970

1980

1

lnAGX

1990

2000

−2 0

.5

5 DlnAGX

10

−2.5 1970 1

1980

1990

2000

−1.75

0

−.1 1970 1

10

−1.5

.1 0

5 lnGLO

DlnAGX

−2 1980

1990

2000

5

lnGLO

10

DlnGLO

1970 1

1980

1990

2000

DlnGLO

.2 0

0

0

5

10

1970

1980

1990

2000

5

10

Figure 7.21. Plot of Variables and their Correlograms Note: The graphs are produced by using the econometric software PCFiML version 9.0 (Doornik and Hendry, 1997).

non-stationary variable the correlograms die down only slowly whilst for a stationary variable they damp down very quickly (just on the ﬁrst lag) and then give random movement. Figure 7.21 shows that the correlograms for lnMAR and lnAGX behave like that of the non-stationary variables while those of DlnMAR and DlnAGX give random movement as expected for stationary variables. The correlogram of lnGLO is most complex: it dies down showing a pattern of positive autocorrelation and then portrays a negative autocorrelation. The correlogram of DlnGLO reﬂects a pattern of stationary variables. Thus correlograms cannot conﬁrm the order of integration of lnGLO; it can be either an I(1) or an I(0) variable. 7.4.4.3. ESTIMATION STRATEGY Once it is determined that variables in the model have integrating properties, the only way to infer about the long-run relationship is to employ some kind of cointegration technique. There are several cointegration methodologies—the at lag k divided by variance, i.e. pk ¼ gk=g0. If pk is plotted against k, the graph we obtain is known as a correlogram.

217

The Implications of Declining Commodity Prices simplest one is the Engle-Granger two-step procedure. The basic idea behind Engle-Granger technique is that if two variables, say Yt and Xt are both ~I(d), a linear combination of them, such that Vt ¼ Xt uYt , in general will also be ~I(d). Engle and Granger, however, showed that in an exceptional case if the constant u yields an outcome where Vt I(d-a) and a>0, then Xt and Yt will be cointegrated. Thus if Xt and Yt are ~I(1) they will be cointegrated and have a valid long-run relationship if residuals from the OLS regression of Xt on Yt is ~I(0). This is the ﬁrst step in the Engle-Granger procedure. On the other hand, if variables are cointegrated there will exist an error-correction model (ECM) of that cointegrating relationship, which gives the short-run dynamics in the second step. Assuming that both Yt and Xt are ~I(1) so that DYt and DXt are ~I(0), the short-run error correction model (ECM) can be represented as:

D ln Yt ¼ p0 þ

m X i¼0

pli D ln Xt þ

n X

p2i D ln Yt þ p3 y^t1 þ j

(5)

i¼1

where, y^t1 is the lagged error from the cointegrating relationship and j is the white noise. It is worth noting that the ECM is not a mere regression of the stationary variables; rather it includes y^t1 , the deviation from the long-run relationship. Thus the ECM captures the short-run deviations taking long-run information into account. A valid representation of the ECM will require 0 > p3 $ 1. The usual practice with error-correction modelling is to follow the ‘general to speciﬁc’ methodology by constructing a general model in the beginning and subsequently reducing it to a parsimonious form after dropping all the insigniﬁcant variables step-by-step. Thus one could employ the Engle–Granger cointegration procedure to estimate equation (2) and test for a valid long-run relationship. However, since the ﬁrst step of the Engle-Granger procedure is basically an OLS regression involving non-stationary variables, it yields standard errors that do not provide the basis for valid inferences. Thus in estimating equation 2, we cannot be certain whether each of the explanatory variables are individually statistically signiﬁcant even when the equation turns out to be a cointegrating relationship.43 We propose to handle this problem by using the Phillips-Hansen Fully Modiﬁed OLS (PHFMOLS) technique (Phillips and Hansen, 1990). The PHFMOLS is a method of optimal single equation technique, which is asymptotically equivalent to the maximum likelihood procedure. It makes a semi parametric correction to the OLS estimator to eliminate the dependency on the nuisance parameters and provide standard errors that follow a standard normal distribution asymptotically and thus are valid for drawing statistical inferences. Due to its advantages the 43 It might be possible that only one of the explanatory variables is signiﬁcant resulting in a cointegrating relationship while the other right hand side variable does not have any inﬂuence on the model.

218

Marginalization and World Trade use of PHFMOLS has become quite popular in international trade and macroeconometric modelling.44 Yet another problem arising from the estimation of the long-run relationship is that unit root tests of the variables could not conﬁrm the non-stationarity of the lnGLO variable. If lnGLO is indeed stationary on its level, there will be a mixture of I(1) and I(0) variables on the right hand side of the model posing the question whether an I(0) regressor plays a role in determining a dependent variable which is ~I(1). Holden and Perman (1994) considered a model with two I(1) and an I(0) variables. The authors tested for the long-run relationship between the two I(1) variables and included the I(0) variable only in the short-run error-correction model. This procedure thus assumes that the I(0) variable does not play a role in the long-run behaviour of the model even disregarding the economic theory behind it. In contrast, Pesaran et al. (2001) strongly argued that the fact that the variables in the estimating equation have different orders of integration does not necessarily mean that they are unlikely to have any long-run impact. Pesaran et al. also devised a strategy, which tests the existence of a long-run relationship when the regressors are a mixture of I(0) and I(1) variables. For this paper we will use this test to determine the long-run relationship in equation (2). 7.4.4.4. TEST FOR EXISTENCE OF A LONG-RUN RELATIONSHIP First, we apply the Pesaran et al. test to ascertain whether the model in (2) is a cointegrating relationship. This test is based on an OLS estimation of an unrestricted error-correction model, a general speciﬁcation of which with respect to our model can be written as: ln MARt ¼a þ F1 ln MARt1 þ F2 ln AGXt1 þ F3 ln GLOt1 þ

p X i¼1

pi D ln MARt1 þ

g X i¼0

p2 D ln AGXti þ

g X

p3 D ln GLOti þ qt

(6)

i¼0

where, all the variables are deﬁned as above and the last term on the right hand side is the white noise. Estimation of (5) in itself is not interesting since the existence of a long-run relationship can only be tested by examining the joint null hypothesis that F1 ¼ F2 ¼ F3 ¼ 0 with the help of either a Wald or an F-test. The presence of a long-run relationship requires the rejection of this null. However, the asymptotic distribution of these test statistics is non-standard and Pesaran et al. provide the necessary critical upper (FU ) and lower bound (FL ) values for the tests.45 The FU -statistics are derived under the assumption that all 44 Among others Athukorala and Riedel (1996), Muscatelli (1995) and Senhadji and Montenegro (1998) have used the same technique for trade modelling while Mallick (1999) is an example of the application of the procedure in macroeconometric modelling. 45 Pesaran et al. give both the critical values for Wald-and F-statistics. In this paper we will only consider the F-statistics.

219

The Implications of Declining Commodity Prices variables are ~I(1) and FL consider all of them to be ~I(0). If the computed F-statistic (F ), which is obtained by restricting that F1 ¼ F2 ¼ F3 ¼ 0, is greater than the critical upper value, i.e. F > FU , one can reject the null and conclude that there exists a valid long-run relationship between the variables in the equation. If F < FL , there is no long-run relationship and ﬁnally if FL < F < FU the test is inconclusive. Pesaran et al. (p.290) clearly point out ‘[I]f the computed Wald-or F-statistic falls outside the critical value bounds a conclusive inference can be drawn without needing to know whether the underlying regressors are I(1), cointegrated amongst themselves or individually I(0)’. In order to determine the long-run relationship equation (6) was run with p¼1 and g¼0.46 In our case the computed F-test statistic was 5.97 against its critical FU value of 4.85.47 This thus rejects the null hypothesis of non-cointegration and suggests the existence of a valid long-run relationship between the variables, i.e. the share of agriculture exports in world merchandise exports and the measure of globalization do determine the marginalization of the LDCs in world trade. 7.4.4.5. ESTIMATING THE LONG-RUN RELATIONSHIP We now proceed to know the exact nature of the long-run relationship by estimating the model. For reasons discussed earlier, the estimation is done by applying the PHFMOLS procedure, the results of which are given in Table 7.20. The estimated results show that all variables are highly signiﬁcant at the one per cent level. The coefﬁcient on lnAGX is positively signed as expected. Thus over the long-run a one per cent fall in the share of agricultural products in world exports reduces LDCs’ share by 1.36 per cent. The imposition of a unit coefﬁcient on lnAGX resulted in a Wald-statistic of 17.21 against its 95 per cent critical value 3.84, thereby rejecting the restriction. This suggests that a certain percentage fall in agriculture share will have an even greater impact on the marginalization of the LDCs. The sign on lnGLO is also negative providing support Table 7.20. PHFMOLS Estimates of the Model Regressor

Coefﬁcient

Standard Error

t-ratio

Constant lnAGX lnGLO R2

3.34*** 1.36*** 0.68***

0.34 0.08 0.15 0.91

9.82 15.53 4.61

Note : Statistical signiﬁcance at the one per cent level is in ***.

46 Since we have a small sample over-parameterization of the model can be very problematic in terms of having fewer degrees of freedom. Such choice of lag length can be rationalized by the use of annual data. 47 This critical value is based on an unrestricted intercept and no trend as reported in Table C1.iii in Pesaran et al. (2001).

220

Marginalization and World Trade for our hypothesis. A one per cent rise in the world export-GDP ratio reduces the relative importance of LDCs by 0.68 per cent. The long-run model can explain the 91 per cent variation in the trends in marginalization of the LDCs. 7.4.4.6. SHORT-RUN DYNAMICS The existence of a long-run cointegrating relationship would imply a short-run error correction model, which we model under the framework of error correction modelling strategy. The error-correction model regresses the current value of the dependent variables in stationary form onto its own lagged value, current and lagged values of the stationary form of the independent variables and the lagged error term from the cointegrating equation. The general to speciﬁc methodology is used to ﬁnd a parsimonious representation of the relationship. In initial experiments the model was estimated by taking ﬁrst order lag of the ﬁrst differences of the dependent and independent variables and including the lag of the longrun errors (ECMt1 ). Then most insigniﬁcant variables were deleted one by one to give the most parsimonious representation of the short-run model. It is to be mentioned that initial runs were confronted by non-normality of errors, which could be detected due to a large unexplained movement in the dependent variable for 1996. A dummy variable for 1996 is thus inserted to overcome the problem. It is now evident from Table 7.21 that in the short-run only the indicator of globalization has a signiﬁcant inﬂuence on marginalization of the LDCs. The sign on lnGLO is also negative suggesting that even in the short-run LDCs cannot take the advantage of global integration and a rise in the level of globalization depresses their relative importance in world exports of merchandise goods. Interestingly, however, the short-run model fails to ﬁnd a signiﬁcant effect of share of agriculture on the marginalization of the LDCs. The reason might be that in the short-run potentially there may be many other variables that are likely to affect the export performance of the LDCs, which have not been modelled here. This is also reﬂected in the somewhat lower size of the explanatory power of the model as the adjusted R2 turns out to be only 0.48. The error-correction term is correctly signed and signiﬁcant implying that the long-run model is correctly speciﬁed as the short-run model converges to the long-run relationship. However, the speed of convergence is slow as only 23 per cent of adjustments are corrected within a year. This slow adjustment mechanism reveals that exogenous shocks (may be natural calamities or other policy factors) in some or in all countries, which cause the share of LDCs as a group to fall will have a considerably long lasting effect before the trends in marginalization can again be explained by the share of agriculture in total global exports and world exports-GDP ratio. Diagnostic tests for the short-run model did not report any problem, as there was no evidence of residual serial correlation, non-normality, heteroscedasticity and functional form problem associated with the model at the 95 per cent level of conﬁdence.

221

The Implications of Declining Commodity Prices Table 7.21. Short-Run Error Correction Model DlnMAR ¼

0.49***

0.35***

(S.E.) t-ratio

(0.01) 4.78

(0.12) 2.95

DlnGLOt1

þ 0.19*** (0.05) 3.39

Adjusted R2 ¼ 0.48 Serial Correlation [x2 (1)] ¼ 0:12 Normality [x2 (2)] ¼ 0:82

D96 0.23** ECMt1 (0.08) 2.67 F(3, 24) ¼ 9.27*** Functional Form [x2 (1)] ¼ 0:19 Heteroscedasticity [x2 (1)] ¼ 0:08

Note : *** and ** are for statistical signiﬁcance at the one and ﬁve per cent level respectively. D96 is the dummy variable representing 1 for 1996 and 0 for all other years. For diagnostics Godfrey’s (1978) LM test for serial correlation, Ramsey’s (1969) RESET test for functional form, Jarque-Bera (1987) test for normality of residuals and White’s (1980) test for heteroscedasticity are performed. The critical values for x2 (1) and x2 (2) at the 95 per cent level are 3.84 and 5.99 respectively, which are being used to test the null hypotheses of no serial correlation, no functional form problem, normality of regression residuals and homoscedastic errors. Since in every case the computed statistics are smaller than the corresponding critical values, all the null-hypotheses are maintained or, in other words, the model passes all the diagnostic tests.

7.4.5. Estimation of the model for small states 7.4.5.1. BIVARIATE RELATIONSHIP Estimation for small states will follow the same procedures as described in the case of LDCs. First, we consider the bivariate relationship between marginalization of small states (MARSS), measured by the combined share of small states in world merchandise exports, and share of agricultural exports in world exports (AGX) and between MARSS and globalization, measured by world export-GDP ratio. Figure 7.22 exhibits the scatter plots of logged MARSS (lnMARSS) and lnAGX. A fairly strong positive relationship is found as the estimated R2 is 0.74. In contrast, the relationship between lnMARSS and lnGLO, presented in Figure 7.23, is rather weak; the estimated R2 is only about 0.15. However, we need to estimate the model formally and test for existence of a long-run relationship. 7.4.5.2. TESTS FOR UNIT ROOTS AND COINTEGRATION Since the units roots for lnAGX and lnGLO have already been tested above, such tests are required for only lnMARSS. The DF and ADF regressions for lnMARSS resulted in the following test statistics. Thus it is observed that the DF and ADF test statistics for lnMARSS, both with and without the trend term, are absolutely smaller than the 95 per cent critical values suggesting that lnMARSS contains a unit root on its level. However, when the same tests were performed on the ﬁrst difference of lnMARSS, DlnMARSS, the null hypothesis of unit root was overwhelmingly rejected at the 95 per cent level. Hence, it can be concluded that lnMARSS is ~I(1). We recall that lnAGX was found to be ~I(1) as well, while lnGLO turned out to be ~I(0). Since the variables are a mixture of I(1) and I(0) variables, the Pesaran et al. test has been carried out to test for the existence of a long-run relationship.

222

Marginalization and World Trade −2.7

−2.5

−2.3

−2.1

−1.9

−1.7

−1.5 −0.6

lnSMARSS = 0.5173 lnAGX + 0.0005 R 2 = 0.7414

−0.7 −0.8

lnMARSS

−0.9 −1 −1.1 −1.2 −1.3 −1.4 −1.5 lnAGX

Figure 7.22. Scatter Plot of lnMARSS and lnAGX for Small States −2.3

−2.2

−2.1

−2

−1.9

−1.8

−1.7

−1.6

lnMARSS = −0.3538 lnGLO −1.7253 R 2 = 0.1464

−1.5

−1.4 −0.6 −0.7 −0.8

lnMARSS

−0.9 −1 −1.1 −1.2 −1.3 −1.4 −1.5 lnGLO

Figure 7.23. Scatter Plot of lnMARSS and lnGLO for Small States

Initial experiments suggested signiﬁcant unexplained movement in residuals for 1974 and 1993 and thus the existence of a long-run relationship was tested including two dummy variables for those two atypical years.48 The F-test statistic arising out of the Pesaran et al. test was estimated to be 5.01 against its critical upper (FU ) value of 4.85. Since the computed F-statistic exceeds the critical F value we can conclude that the variables are cointegrated and consequently there is a valid long-run relationship between the marginalization of small states, share of agricultural exports in world exports and world export-GDP ratio, as hypothesized in our model. 48 If the dummies are not inserted, the graphical plot of the residuals was found to be non-normal.

223

The Implications of Declining Commodity Prices Table 7.22. Unit Root Test for lnMARSS DF-ADF tests without the trend term Variable inMARSS DlnMARSS

DF-ADF tests with the trend terms

DF

ADF

DF

ADF

0.59 5.34*

0.51 4.23*

1.87 5.25*

1.81 4.13*

Note : D implies ﬁrst difference. The 95 per cent critical values for DF and ADF tests with and without the trend term are respectively 2.97 and 3.57. * indicates rejection of the null hypothesis of unit root at the 95 per cent level.

7.4.5.3. THE LONG- AND SHORT-RUN RELATIONSHIPS Since the long-run relationship was tested by inserting two dummy variables, their inclusion may be justiﬁed if they are found to be statistically signiﬁcant in the long-run model. As for reasons discussed earlier, the long-run relationship is estimated by the procedure of the Phillips-Hansen Fully Modiﬁed OLS (PHFMOLS). The long-run model thus estimated is given below: lnMARSS ¼ 5.15*** þ 0.42*** lnAGX 0.18** lnGLO þ 0.17*** D74 þ 0.21*** D93 (s.e.) (0.17) (0.04) (0.07) (0.04) (0.05) t-ratio 30.9 9.89 (2.57) (4.17) (3.91) R2 ¼ 0:85

Therefore, in the long-run agriculture share is positively associated with the marginalization of small states in contrast to an inverse relationship between the dependent variable and globalization. A one percentage point fall in AGX results in 0.42 per cent fall in the share of small states over the long-run, while a rise in GLO by the same magnitude will result in a decline of small states’ share in world merchandise exports by 0.18 per cent. AGX and GLO are statistically signiﬁcant respectively at the one and ﬁve per cent levels. Both the dummies (D74 and D93) are also highly signiﬁcant at less than one per cent level justifying their inclusion in the model. The long-run model explains 85 per cent variation in lnMARSS. The short-run dynamics are modelled following the error-correction methodology. In the short run, no effect of agriculture share is observed although a negative relationship between globalization and marginalization of small states is maintained. The dummy variables for 1974 and 1993 are also signiﬁcant in the short run relationship. The error-correction term (ECMt1 ) is correctly signed and signiﬁcant at the ﬁve per cent level and reveals that 44 per cent disequilibrium errors are corrected within a year. The model, however, explains only 38 per cent variation in the dependent variables suggesting that there are other factors explaining most of the marginalization in the short run. The diagnostic test statistics concerning the null-hypotheses of no serial correlation, no functional form problem, non-normality and homoscedasticity of errors are maintained by the statistical tests.

224

Marginalization and World Trade Table 7.23. Short-Run Error Correction Model DlnMARSS ¼ 0.28** 0.25** DlnMARSSt1 DlnGLOt1 D93 0.15*** D74 0.44**ECMt1 0.25** þ0.16*** (S.E.) t-ratio

(0.01) 2.78

(0.10) 2.55

2.02

Adjusted R2 ¼ 0:38 Serial Correlation [x2 (1)] ¼ 0:38 Normality [x2 (2)] ¼ 1:16

(0.12) 3.20

(0.05) 3.78

(0.04) 2.05

(0.22)

F(5, 20) ¼ 7.67*** Functional Form [x2 (1)] ¼ 0:75 Heteroscedasticity [x2 (1)] ¼ 0:25

Note : *** and ** are for statistical signiﬁcance at the one and ﬁve per cent levels respectively. For diagnostics Godfrey’s (1978) LM test for serial correlation, Ramsey’s (1969) RESET test for functional form, Jarque-Bera (1987) test for normality of residuals and White’s (1980) test for heteroscedasticity are performed. The critical values for x2 (1) and x2 (2) at the 95 per cent level are 3.84 and 5.99 respectively, which are being used to test the nullhypotheses of no serial correlation, no functional form problem, normality of regression residuals and homoscedastic errors. Since in every case the computed statistics are smaller than the corresponding critical values, all the null-hypotheses are maintained or, in other words, the model passes all the diagnostic tests.

7.4.6. Conclusion of results In this section an attempt was made to explain the trends in the share of exports of LDCs and small states in terms of agriculture-exports and exports-GDP ratios in the world economy. In particular, we examined whether there was a valid long-run relationship among the variables, as speciﬁed in the model. In light of the problems associated with the time series properties of the variables, which might lead to a spurious relationship, careful attention was given by testing the variables for unit roots and using appropriate cointegration methodology. It was found that the variables were integrated on their levels; nevertheless a genuine long-run relationship among the variables in the model, both for LDCs and small states, was conﬁrmed. In the long run, as the share of agriculture in total exports falls and the ratio of world exports to GDP rises, LDCs’ and small states’ relative importance shrinks. These two variables together can explain 91 and 85 per cent variation in marginalization trends respectively for LDCs and small states. The short-run dynamics were modelled following the error-correction methodology where only globalization was found to affect relative importance of the two country groups negatively. The short-run models explained a relatively small variation in the dependent variable and it is possible that other exogenous and policy factors might have contributed to the declining importance of LDCs and small states in the short run.

7.5. Implication for Long-term Trade and Development While the dependence on primary products and increasing globalization in the world economy can explain much of the general trend in declining relative signiﬁcance of LDCs and small states, there are other factors that aggravate the process, either by inhibiting or by not facilitating the development of dynamic

225

The Implications of Declining Commodity Prices export sectors. The long-term trade and development prospects of LDCs and small states hinge critically upon the interplay of these factors and without addressing them the process of marginalization cannot be checked. To conclude this paper therefore we provide brief discussions on these issues below. First, the existing structure of export trade does not allow LDCs and small states to take full advantage of high income growth in the world economy. One straightforward policy recommendation would be diversiﬁcation of their export basket by aiming at production and export of manufactured goods. This optionhas, however, so far proved to be a very difﬁcultone. Although most of these countries have a natural comparative advantage in the production of primary products, in the past many of them pursued an inward-looking import-substitution strategy in order to facilitate the formation of a manufacturing industrial base in the domestic economy. The import-substituting industries that were developed under the protective regime remained inefﬁcient and, in the face of severe external and internal imbalances affecting the countries, a policy for trade liberalization and reforms were carried out. Since import-substitution regimes resulted in policy-induced biases against agriculture, a policy reversal to exportpromotion strategy only revived the static comparative advantage of primary commodities. Thus the export structure continues to be dominated by primary commodities, thereby leaving the process of marginalization uninterrupted. Second, due to the small size of the domestic market and low purchasing power of consumers an efﬁcient manufacturing industrial base in LDCs and small states can only ﬂourish if they can engage in international trade.49 Most small states and quite a few LDCs are, however, confronted by natural barriers to trade associated with unfavourable geographical characteristics (such as remoteness and isolation), which increase costs of both export and import trade relative to countries with more favourable geographical characteristics. In particular, small states pay higher transportation costs because of geographical locations, small volume of cargo, bulky low-value products (e.g. agricultural commodities) and lack of equivalent return cargo. The ﬁgures quoted in Bernall (2001) show that transportation and freight costs for some small states are as high as 30 per cent of export volume compared to only 4 per cent for large states. Similarly, sub-Saharan Africa’s net insurance freight costs account for 15 per cent of their total exports as against 5.8 per cent for all developing countries (Amjadi and Yeats, 1995).50 When export structure is characterized by a high share of bulky low-value products (e.g. agricultural commodities), countries face much higher freight costs than high-value products with low storage factors (e.g. many manufacturing exports). 49 The small size of the domestic market does not allow ﬁrms to exploit either internal economies of scale (i.e. where unit cost is reduced as the size of the ﬁrm gets bigger) or external economies of scale (i.e. where unit cost is inﬂuenced by the size of the industry). 50 According to the World Bank (1996) net transport and insurance payments average more than 25 per cent of total exports for one-third of sub-Saharan African countries.

226

Marginalization and World Trade These excessive costs alone can serve to make the poor and vulnerable countries’ exports uncompetitive. There is some evidence that increase in transport costs reduces trade volumes (Limao and Venables, 2001) and an ad valorem transport cost of 20 per cent on both ﬁnal output and intermediate goods reduce the domestic value added (and thus GDP) by 60 per cent when intermediate goods account for 50 per cent of costs (Redding and Venables, 2001).51 As excessive transport costs substantially reduce the domestic value added out of the production of export goods dependent on imported inputs, they not only affect international competitiveness but also discourages foreign ﬁrms to relocate their production to these countries even when the wages are low. Third, most small states and LDCs also suffer from a poor state of physical and social infrastructure (human capital), the development of which is considered to be vital for expanding productive capacities and particularly for exporting the manufactured goods that have witnessed rapid growth in world trade.52 However infrastructure development is very expensive and requires long-term investment. Given the current level of income and domestic savings the development of infrastructure and the level of domestic investment in many poor countries will critically depend on the inﬂow of ofﬁcial development assistance (ODA).53 The data in Table 7.24 show that during the late 1990s the ﬂow of ODA to developing countries, LDCs, and small states declined absolutely. However, while developing countries, on the whole, managed to enjoy increased total ﬁnancial ﬂows from about $74 billion in 1990 to about $79 Table 7.24. Ofﬁcial Financial Flows ($ million) Total ﬂows

Total ODA

Year

Developing Countries

LDCs

Small States

Developing Countries

LDCs

Small States

1975 1980 1985 1990 1995 1999

21905 42591 41019 74122 70725 79165

4489 9872 10257 17470 17064 11797

982 1693 1730 2872 1792 950

16142 32460 30180 56036 58706 50543

3713 8724 9483 16747 17198 11591

865 1505 1353 2427 1811 1076

Source : UNCTAD (2001).

51 This is compared to a country that faces zero transport costs. Redding and Venables revealed that more than 70 per cent variation in cross-country per capita income could be explained by the geography of access to markets and sources of supply of intermediate inputs. 52 Since most manufacturing exports require a relatively higher input of capital and skill per worker than land per worker, Wood and Mayer (2001) argue that Africa, which includes a number of LDCs, does not have a comparative advantage in exporting labour-intensive manufacturing because of its higher endowment of natural resources (land) to human capital. 53 Gross domestic investment as the percentage of GDP in LDCs has more or less remained unchanged in the 1990s: 22.7 per cent on an average in 1990–94 as against 23.3 per cent in 1997. With such low levels of investment LDCs can hardly create new productive capacities after replacing the depreciation or stock destroyed by factors such as civil war or sheer neglect (UNCTAD, 1999).

227

The Implications of Declining Commodity Prices 3

Per cent of global FDI inflow

2.5 LDCs

Small States

2

1.5

1

0.5

2000

1998

1996

1994

1992

1990

1988

1986

1984

1982

1980

1978

1976

1974

1972

0

Figure 7.24. Share of LDCs and Small States in Global Inﬂow of FDI Source: Authors’ estimates from UNCTAD (2002b).

billion in 1999, LDCs and small states saw the ﬂow going down by respectively about 34 and 67 per cent.54 Perhaps the most obvious and important source of ﬁnancing domestic investment is the inﬂow of foreign direct investment (FDI). In 2000 total inﬂow of FDI in the global economy stood at $1271 billion but for LDCs and small states the ﬁgures were only $4.4 and $3.7 billion respectively. During 1990–2000 world FDI inﬂow grew at a trend rate of 14.2 per cent whereas the comparable rates for LDCs and small states were computed to be 11.6 and 9.57 per cent respectively.55 Figure 7.24 shows that while small states’ share of FDI has fallen from a high of about 2.5 per cent of total FDI in 1972 to less than 0.5 per cent, LDCs are down from a peak of over 1.5 per cent in the late 1970s to less than 0.5 per cent by the end of 1990s, both reﬂecting a clear negative trend. Fourth, for a long time LDCs and small states have beneﬁted from various preferential trading arrangements. The evolving trading system, however, either has reduced the preferential trade margins for these countries or threatens to erode the preferences altogether. For example, in the post-Uruguay Round 54 In the 1990s total ODA contributions from OECD donor countries allocated to LDCs fell by 29 per cent (UNCTAD, 1999) and aid per capita to the developing countries as a whole declined by nearly a third from $32.27 to $22.41 (Stiglitz, 2000). 55 For the period of 1990–2000 the trend growth rates of FDI inﬂow into developed and developing countries are estimated to be respectively 13.63 and 15.10 per cent.

228

Marginalization and World Trade period average tariffs on industrial goods stood at only 3.9 percent providing a very low margin of preference to the recipient countries. Again, under the Lome´ convention many small states have enjoyed preferential trade margins extended by the EU, which have become incompatible under the WTO regime.56 WTO compatibility of these provisions will require substantial opening-up of the sectors currently protected for the beneﬁciary countries.57 As a consequence, the net economic effect of the Uruguay Round trade liberalization upon the highly trade preference dependent economies has been found to be negative (Grynberg, 2001). Therefore, it appears that the global trading regime under the WTO will have further consequences on exports and trade of small states and LDCs. Last but not least, factors associated with internal or domestic economy in many LDCs and small states have adversely affected their export trade. Improper interventions resulting in inefﬁciencies and leading to wastage of resources, social and political unrest creating a domestic environment hostile to investment and production, inefﬁcient and lengthy bureaucratic procedures together with corruption causing high transaction costs, all combine to make the costs of doing business very high; this in turn reduces the competitiveness of tradable activities. Nowhere are social and political stability and good governance more important than in LDCs and in small states that suffer from structural obstacles such as highly concentrated export structure and unfavourable geographical location. It goes without saying that overall competitiveness and export success of these economies in future will critically hinge to a great extent upon the formation of an efﬁcient administrative and institutional framework. Small states and LDCs pose a challenge to the international community in the ongoing process of globalization. Increased integration and rising trade and investment in the world economy may not beneﬁt them substantially. The problem of marginalization in world trade is mostly associated with their inability to diversify exports. Most small states and LDCs have static internal comparative advantage in primary activities, the relative importance of which have shrunk considerably in world trade over the past few decades. Thus, an emphasis on static comparative advantage in the allocation of resources might act as a hindrance to diversiﬁcation of export structure. One move in the right direction might be

56

Grynberg (2001) provides a detailed discussion on this. Some small states are heavily dependent on various commodity arrangements with the EU. These typically cover exports from small states that would not be competitive in the world market, but are of major economic and social importance to these countries. For example, the Sugar Protocol offers valuable protection to St Kitts and Nevis, where sugar revenue accounts for about 50 per cent of GDP. In light of the problem of WTO incompatibility these arrangements are being renegotiated and revised resulting in considerable erosion of trade preferences to small states (Berthelot, 2001). 57

229

The Implications of Declining Commodity Prices to introduce new products, where possible, in the form of processed primary products or labour intensive light manufacturing exports. It needs to be stressed that although small states can achieve efﬁciency gains, trade liberalization itself will not remove non-policy barriers to trade imposed by unfavourable geographical locations and undeveloped human and physical infrastructure. While little can be done to overcome geographical locational disadvantages, given the low level of domestic saving and poor capacity of internal resource mobilization for development of infrastructure, an increased inﬂow of ODA is indispensable. Finally, small states and LDCs should also look into their own domestic economies. Civil unrest, poor law and order, and inefﬁcient, corrupt and lengthy bureaucratic procedures in a large number of LDCs and small states undermine the environment for production and business activities. Only concerted efforts at the domestic fronts of these countries combined with cooperation extended by the international community can help mitigate the problem.

230

Marginalization and World Trade Appendix 7.1. List of LDCs Sl. No.

Country

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49

Afghanistan Angola Bangladesh Benin Bhutan Burkina Faso Burundi Cambodia Cape Verde Central African Republic Chad Comoros Congo, DR Djibouti Equatorial Guinea Eritrea Ethiopia Gambia Guinea Guinea-Bissau Haiti Kiribati Lao People’s Dem. Rep. Lesotho Liberia Madagascar Malawi Maldives Mali Mauritania Mozambique Myanmar Nepal Niger Rwanda Samoa Sao Tome and Principe Senegal Sierra Leone Solomon Islands Somalia Sudan Tanzania Togo Tuvalu Uganda Vanuatu Yemen Zambia

Remark

Commonwealth Member?

Oil Exporter YES

Small State

Small State Small State Small State Data not available Small State

YES

Small State Small State

YES

Small State

YES YES

YES

Small State Small State

Small State

YES

YES YES

YES Data not available Small State

YES YES YES YES

Note : The list of LDCs is from UNCTAD (2002b).

231

The Implications of Declining Commodity Prices Appendix 7.2. List of small states Sl. No.

Country

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

Antigua and Barbuda Bahamas, The Bahrain Barbados Belize Botswana Cape Verde Comoros Cyprus Djibouti Dominica Equatorial Guinea Fiji Gabon Gambia Grenada Guyana Jamaica Kiribati Lesotho Maldives Malta Mauritius Namibia Papua New Guinea Samoa Sao Tome and Principe Seychelles Solomon Islands St Kitts and Nevis St Lucia St Vincent and Grenadines Suriname Swaziland Tonga Trinidad and Tobago Vanuatu

232

Remark

Data Problem OIL

Commonwealth Member? YES YES YES YES YES

LDC LDC LDC YES LDC YES OIL LDC

LDC LDC LDC

Data Problem LDC LDC LDC

OIL LDC

YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES YES

Marginalization and World Trade Appendix 7.3. Merchandise and Services Exports of Individual LDCs

Merchandise Exports ($ mill) Countries Afghanistan Angola Bangladesh Benin Bhutan Burkina Faso Burundi Cambodia Central Af. Rep. Chad Comoros Congo, DR Eritrea Ethiopia Guinea GuineaBissau Haiti Lao PDR Liberia Madagascar Malawi Mali Mauritania Mozambique Myanmar Nepal Niger Rwanda Senegal Sierra Leone Somalia Sudan Tanzania Togo Tuvalu Uganda Yemen Rep. Zambia

90–94 avg. 1995–2000 avg. 2000

Merchandise share in total exports Exports of Commercial services ($ mill) (per cent) 90–94 avg. 1995–2000 avg.

2000 1995–2000 avg. 2000

328.4 3381.8 2007.9 347.0 70.6 99.5 84.3 267.6 107.0

140.8 3888.2 4646.2 430.7 113.0 257.3 66.8 740.8 164.9

125.0 6646.0 6399.0 376.0 140.0 228.0 50.0 770.0 167.0

4.8 103.2 377.4 117.4 n.a. 38.4 7.0 48.0 16.6

7.2 151.0 293.2 136.8 n.a. 43.3 3.0 130.3 11.5

8.0 150.0 283.0 155.0 n.a. 46.0 2.0 159.0 10.0

95.2 96.3 94.1 75.9 n.a. 85.6 95.7 85.0 93.5

94.0 97.8 95.8 70.8 n.a. 83.2 96.2 82.9 94.4

168.7 19.6 1044.0 n.a. 245.3 629.0 18.8

227.3 9.6 1702.8 n.a. 492.1 845.8 40.2

182.7 12.0 2887.0 n.a. 508.0 940.0 62.0

18.8 15.2 145.2 n.a. 247.0 6.2 6.4

29.7 36.5 132.8 n.a. 345.5 6.7 4.7

25.0 46.0 137.0 n.a. 387.0 10.0 6.0

88.5 20.7 92.8 n.a. 58.8 99.2 89.6

88.0 20.7 95.5 n.a. 56.8 98.9 91.2

112.5 169.9 334.0 306.3 388.9 365.3 384.2 143.2 532.0 314.9 287.0 70.7 726.8 133.1 127.0 387.6 411.8 252.1 n.a. 216.5 703.0 980.1

142.4 331.4 480.3 270.7 470.5 512.2 460.2 222.4 1007.4 502.8 304.2 62.1 960.5 22.0 130.8 716.7 669.0 418.5 n.a. 523.7 2543.6 902.5

163.6 315.0 500.0 250.0 445.0 510.0 355.0 235.0 1391.0 804.5 320.0 52.9 927.0 13.0 110.0 1155.0 663.2 427.0 n.a. 520.0 4200.0 759.0

n.a. 35.4 41.0 146.6 31.4 60.8 15.0 157.2 147.8 277.6 17.4 24.4 313.0 52.6 n.a. 77.8 241.2 80.2 n.a. 42.8 106.0 75.0

n.a. 92.2 46.0 263.8 35.3 65.3 24.7 280.0 481.0 560.5 12.3 27.7 345.7 76.3 n.a. 44.8 568.5 69.2 n.a. 159.7 148.5 77.3

n.a. 111.0 49.0 314.0 31.0 68.0 29.0 325.0 509.0 410.0 13.0 39.0 349.0 86.0 n.a. 24.0 615.0 52.0 n.a. 186.0 161.0 80.0

n.a. 78.2 91.3 50.6 93.0 88.7 94.9 44.3 67.7 47.3 96.1 69.2 73.5 22.4 n.a. 94.1 54.1 85.8 n.a. 76.6 94.5 92.1

n.a. 73.9 91.1 44.3 93.5 88.2 92.4 42.0 73.2 66.2 96.1 57.6 72.6 13.2 n.a. 98.0 51.9 89.1 n.a. 73.7 96.3 90.5

233

The Implications of Declining Commodity Prices Appendix 7.4. Merchandise and Services Exports of Individual Small States

Merchandise Exports ($ mill) Countries Antigua and Barbuda Bahrain Barbados Belize Botswana Cape Verde Comoros Cyprus Djibouti Dominica Equatorial Guinea Fiji Gabon Gambia Grenada Guyana Jamaica Kiribati Lesotho Maldives Malta Mauritius Papua New Guinea Samoa Sao Tome and Principe Seychelles Solomon Islands St Kitts and Nevis St Lucia St Vincent and Grens. Suriname Swaziland Tonga Trinidad and Tobago Vanuatu

234

90–94 avg. 1995–2000 avg. 2000

Merchandise share in total exports Exports of Commercial services ($ mill) (per cent) 90–94 avg. 1995–2000 avg.

2000

1995–2000 avg. 2000

48.3

40.2

38.2

342.6

395.5

406.0

9.2

8.6

3616.1 196.2 113.8 1795.9 5.1 19.6 945.0 19.0 51.7 69.8

4384.9 265.0 161.8 2393.7 12.0 11.1 1147.3 20.7 53.3 354.6

5700.5 272.4 194.3 2250.0 11.0 12.0 953.5 20.0 53.0 425.0

556.8 652.0 101.0 177.2 32.4 15.2 1997.6 33.6 41.6 6.2

715.0 954.2 130.2 251.3 78.3 36.5 2725.5 27.2 78.8 6.7

830.0 1055.0 152.0 353.0 99.0 46.0 2930.0 28.0 86.0 10.0

86.0 21.7 55.4 90.5 13.3 23.2 29.6 43.2 40.3 98.2

87.3 20.5 56.1 86.4 10.0 20.7 24.6 41.7 38.1 97.7

482.6 2234.7 45.6 22.8 333.1 1118.8 4.8 102.8 45.3 1370.1 1267.2 1956.2

613.9 3127.9 15.9 23.2 532.2 1340.4 6.5 181.5 66.0 1904.2 1618.5 2189.8

596.0 3883.0 9.0 23.0 570.0 1296.1 8.0 180.0 75.9 2336.5 1580.0 2096.0

419.8 260.6 65.4 79.2 100.8 1129.8 13.0 32.0 142.8 846.2 552.6 267.2

540.0 210.8 91.8 118.8 143.8 1749.0 19.2 43.7 306.5 1101.3 937.5 332.7

557.0 221.0 116.0 145.0 166.0 1988.0 21.0 36.0 345.0 1086.0 1067.0 280.0

53.2 93.7 14.8 16.4 78.7 43.4 25.3 80.6 17.7 63.4 63.3 86.8

51.7 94.6 7.2 13.7 77.4 39.5 27.6 83.3 18.0 68.3 59.7 88.2

6.2 14.4

13.9 8.2

14.2 8.0

33.4 4.4

56.7 8.7

61.0 12.0

19.7 48.5

18.9 40.0

51.4 105.6

125.6 145.0

180.0 93.0

177.4 30.0

253.7 51.0

305.0 57.0

33.1 74.0

37.1 62.0

26.0

25.5

30.0

75.2

92.2

93.0

21.7

24.4

117.3 67.2

75.6 46.8

60.0 47.2

190.2 52.6

292.8 103.3

307.0 124.0

20.5 31.2

16.3 27.6

572.2 649.6 13.3 1833.0

488.8 932.7 12.0 2767.2

435.0 881.0 17.0 4044.0

43.0 95.6 12.6 364.6

84.5 106.5 14.8 513.2

85.0 72.0 17.0 628.0

85.3 89.8 44.8 84.4

83.7 92.4 50.0 86.6

21.7

29.9

26.0

63.0

95.2

117.0

23.9

18.2

Appendix 7.5. Growth Rates of Exports of LDCs Services exports growth rate (per cent)

Merchandise exports growth rate (per cent)

Afghanistan Angola Bangladesh Benin Bhutan Burkina Faso Burundi Cambodia Central Af. Rep. Chad Congo, DR Eritrea Ethiopia Guinea Guinea-Bissau Haiti Lao PDR Liberia Madagascar Malawi Mali Mauritania Mozambique Myanmar Nepal Niger Rwanda Senegal Sierra Leone Somalia Sudan Tanzania Togo Tuvalu Uganda Yemen Rep. Zambia

90–94

1995

1996

1997

1998

1999

2000

1995–2000

90–94

1995

1996

1997

1998

1999

2000

1995–2000

7.2 1.4 13.4 13.3 1.1 24.3 9.6 51.6 2.8 2.0 17.2 n.a. 17.4 1.5 48.8 5.8 38.2 4.0 7.4 7.4 8.1 2.3 8.6 30.7 19.5 0.5 11.5 3.1 3.1 21.1 0.4 8.3 3.8 n.a. 24.7 13.6 5.8

0.0 20.7 37.2 5.5 24.2 24.9 1.9 42.5 17.9 17.2 4.5 n.a. 13.4 3.8 6.1 34.1 3.7 14.3 9.6 34.3 31.6 22.6 12.0 6.6 5.2 27.1 31.7 22.5 63.5 6.6 6.1 31.4 136.4 n.a. 8.5 108.2 13.1

0.0 39.9 3.9 25.7 20.7 1.3 62.9 12.3 17.4 32.1 35.2 n.a. 1.2 16.6 32.3 18.2 3.5 28.0 18.5 6.6 0.5 1.9 14.9 12.3 11.6 2.1 11.1 1.8 11.9 3.4 11.7 11.3 49.1 n.a. 27.6 37.5 1.0

3.2 1.7 13.5 19.7 19.2 5.5 120.5 15.0 19.0 2.5 10.3 n.a. 40.8 6.1 133.3 33.3 11.5 2.3 7.0 14.0 27.8 3.2 17.1 16.1 4.7 2.9 46.7 8.2 63.8 0.0 4.2 0.8 78.5 n.a. 5.5 6.4 11.8

6.7 29.2 25.9 2.4 9.3 1.2 24.4 4.3 4.0 10.1 13.2 n.a. 4.4 0.5 44.9 45.8 3.1 6.0 15.5 4.3 0.9 0.0 3.5 23.0 17.9 22.8 31.8 7.0 58.8 13.3 0.3 21.8 12.9 n.a. 9.7 40.3 19.1

14.3 45.6 7.9 1.9 23.4 21.4 16.9 197.0 17.6 22.6 18.3 n.a. 20.0 8.8 88.9 12.0 15.9 6.4 0.9 14.0 2.7 16.7 12.4 5.6 27.0 14.1 0.0 6.1 14.3 23.1 30.9 7.8 0.5 n.a. 3.6 63.1 2.6

8.3 52.9 17.2 10.9 6.1 16.1 7.4 35.4 13.3 9.4 8.2 n.a. 7.3 2.7 21.6 16.3 4.0 0.0 11.6 19.7 9.3 19.6 38.4 44.0 33.6 9.4 11.7 2.9 116.7 10.0 131.7 22.1 7.2 n.a. 11.2 67.2 5.4

2.6 21.3 17.6 0.0 14.0 1.2 1.2 45.5 3.2 4.2 2.7 n.a. 6.0 3.3 26.8 15.1 1.6 6.7 6.6 2.8 8.6 3.1 16.4 13.8 14.9 3.6 7.7 4.4 12.0 2.7 29.4 5.7 28.5 n.a. 2.2 38.2 1.8

59.7 13.4 9.6 9.1 n.a. 6.8 5.5 6.5 2.7 0.8 16.2 n.a. 1.3 16.1 15.9 31.9 45.4 4.7 10.4 4.8 3.9 3.1 15.7 48.7 29.9 15.4 6.9 4.5 69.2 — 23.3 29.7 9.4 n.a. — — 0.8

0.0 24.7 11.9 39.5 n.a. 10.5 33.3 128.9 23.5 43.5 116.4 n.a. 16.5 21.4 0.0 1300.0 21.4 2.2 19.7 13.6 33.3 11.8 26.7 36.7 12.3 50.0 50.0 17.8 17.4 — 86.4 37.7 23.1 n.a. 62.5 35.6 1.3

14.3 100.0 52.5 22.6 n.a. 2.4 0.0 47.6 0.0 3.0 0.0 n.a. 3.5 270.6 50.0 6.1 16.2 0.0 15.5 78.9 2.9 21.1 4.5 19.1 14.7 0.0 18.2 13.5 2.8 — 54.9 6.4 59.4 n.a. 39.4 4.3 1.3

25.0 38.5 19.3 17.1 n.a. 2.3 25.0 1.3 23.1 2.9 0.0 n.a. 0.9 11.1 100.0 65.4 1.3 0.0 4.0 38.2 6.1 8.7 10.3 21.8 17.1 0.0 161.5 4.4 2.7 — 18.9 21.9 23.5 n.a. 13.8 14.3 1.3

16.7 11.5 5.3 24.5 n.a. 2.4 0.0 34.7 0.0 25.7 3.8 n.a. 8.5 5.7 16.7 3.5 48.7 2.2 8.6 14.9 1.6 4.0 2.5 20.3 45.5 0.0 5.9 11.2 2.8 — 53.3 13.6 16.7 n.a. 6.7 20.8 3.9

0.0 26.0 5.6 22.0 n.a. 12.2 33.3 22.4 30.0 3.8 7.9 n.a. 13.6 45.5 20.0 2.2 12.9 9.1 9.8 2.5 9.8 16.7 3.1 19.6 4.8 8.3 15.6 4.1 15.1 — 485.7 16.9 16.9 n.a. 3.4 6.0 8.1

14.3 3.2 6.4 0.0 n.a. 0.0 0.0 32.5 23.1 0.0 0.0 n.a. 1.3 0.0 0.0 12.1 9.9 2.1 8.3 24.4 1.5 3.6 10.2 3.7 9.7 0.0 5.4 0.6 2.4 — 70.7 1.4 3.7 n.a. 2.2 14.2 0.0

3.4 8.0 2.4 7.7 n.a. 3.4 15.3 32.6 6.6 3.3 20.1 n.a. 6.7 42.0 25.6 231.6 13.7 1.1 9.7 11.1 5.7 9.6 9.6 13.7 1.0 9.7 24.1 2.6 0.5 — 62.4 8.5 3.6 n.a. 21.3 8.9 0.9

Appendix 7.6. Growth Rates of Exports by Small States Services exports growth rate (per cent)

Merchandise exports growth rate (per cent) 1996

1997

1998

1999

2000

Antigua and Barbuda 20.4 11.1 43.3 Bahrain 5.9 13.7 14.3 Barbados 0.2 31.3 17.6 Belize 6.6 11.8 8.5 Botswana 0.3 14.1 18.4 Cape Verde 5.2 80.0 44.4 Comoros 5.0 0.0 45.5 Cyprus 4.7 27.0 12.9 Djibouti 6.3 0.0 35.3 Dominica 1.5 4.3 13.3 Equatorial Guinea 15.9 92.1 81.8 Fiji 4.6 12.5 20.8 Gabon 9.0 15.4 21.9 Gambia 10.5 54.3 31.3 Grenada 1.6 8.0 8.7 Guyana 15.8 0.2 13.6 Jamaica 5.0 17.7 3.1 Kiribati 10.7 40.0 28.6 Lesotho 19.0 11.9 16.9 Maldives 1.9 8.9 20.4 Malta 14.4 21.9 9.6 Mauritius 2.3 22.6 1.9 Papua New Guinea 15.5 0.2 5.0 Samoa 19.5 125.0 11.1 Sao Tome and Principe 12.0 37.5 20.0 Seychelles 11.3 2.0 92.3 Solomon Islands 14.3 18.3 3.6 St Kitts and Nevis 5.1 13.6 15.8 St Lucia 0.2 17.0 33.9 St Vincent and Grens. 6.3 16.0 9.5 Suriname 1.3 6.2 9.2 Swaziland 9.4 22.2 6.8 Tonga 10.7 0.0 35.7 Trinidad and Tobago 5.6 26.3 4.1 Tuvalu n.a. n.a. n.a. Vanuatu 3.8 12.0 7.1

23.5 6.7 0.7 3.2 12.1 7.7 50.0 9.9 21.7 3.9 114.5 17.1 8.6 28.6 9.5 24.6 0.0 20.0 4.8 23.7 5.2 3.2 13.9 50.0 25.0 13.0 8.0 18.2 19.5 0.0 61.9 7.6 11.1 0.0 n.a. 16.7

15.4 25.4 11.0 2.5 31.5 28.6 0.0 15.1 17.9 18.9 14.9 17.7 36.5 80.0 17.4 24.8 5.1 16.7 1.0 1.4 11.0 0.0 17.9 0.0 20.0 8.0 28.0 7.7 6.1 8.7 37.8 0.6 20.0 11.9 n.a. 2.9

33.3 26.6 4.8 7.1 35.7 20.0 44.4 6.0 13.0 14.3 68.8 11.0 30.2 74.1 70.4 8.1 5.5 14.3 11.3 13.5 8.7 16.7 8.4 33.3 33.3 18.9 15.9 0.0 4.8 2.0 3.2 7.1 50.0 23.9 n.a. 23.5

0.0 37.7 3.0 16.9 1.7 8.3 40.0 4.3 0.0 1.9 31.1 7.8 34.0 0.0 8.7 9.0 4.5 175.0 22.1 18.8 18.0 19.6 8.9 30.0 25.0 24.1 36.3 7.1 1.7 4.1 3.3 9.6 8.3 52.3 n.a. 3.8

Country

90–94

1995

1995–2000 90–94 12.3 10.0 7.7 7.5 7.9 19.2 0.0 0.7 4.4 2.6 62.2 0.3 9.4 7.6 14.9 5.0 1.4 34.8 7.2 9.9 7.5 3.1 3.2 31.6 11.8 26.4 4.3 5.9 7.6 0.6 3.5 1.2 0.5 15.8 n.a. 0.9

8.8 7.4 3.5 13.2 23.9 12.9 29.4 13.2 5.5 16.7 3.5 8.6 3.0 9.2 13.6 10.0 12.2 24.3 2.9 18.6 9.3 3.1 12.3 7.2 21.7 7.4 16.6 13.2 12.1 10.9 16.9 6.0 7.4 5.0 n.a. 13.4

1995

1996

1997

1998

1999

2000

11.0 4.3 10.7 5.5 1.4 5.6 16.6 2.5 4.4 13.8 3.3 10.8 7.5 6.8 3.6 6.6 0.2 5.8 7.3 6.8 0.8 1.6 15.6 7.8 34.9 38.6 29.0 28.9 43.6 2.0 46.2 14.0 21.5 6.3 29.7 3.1 28.6 7.4 13.8 18.2 15.4 2.2 13.2 3.9 2.5 4.4 9.1 1.1 15.2 3.6 0.0 3.7 3.8 3.7 19.6 9.8 19.4 5.0 13.1 9.5 33.3 25.0 20.0 83.3 1300.0 28.6 7.2 8.0 8.8 23.4 7.0 10.3 4.5 7.3 0.0 5.4 28.4 11.2 50.0 131.6 10.2 2.1 14.1 2.7 2.0 7.1 0.0 10.5 24.1 0.7 11.1 2.3 8.3 0.0 1.4 13.7 8.0 0.4 6.2 4.2 11.8 2.0 11.8 10.5 14.3 0.0 0.0 16.7 0.0 13.3 132.4 41.8 19.6 2.7 17.9 24.3 8.0 6.5 3.3 1.5 5.1 2.6 3.5 6.2 3.6 9.4 11.8 21.1 8.7 4.0 16.7 3.6 36.6 34.6 8.1 19.9 22.0 12.9 32.5 17.0 4.8 1.7 19.0 29.8 20.0 16.7 0.0 0.0 85.7 7.7 7.5 11.0 4.1 15.2 12.0 2.3 16.7 34.3 36.2 18.8 1.9 11.8 13.0 7.5 9.3 9.6 5.8 4.1 11.4 0.8 7.9 9.1 2.2 4.1 18.0 33.3 2.1 8.2 17.0 0.0 9.0 16.4 44.3 5.0 11.5 21.2 36.4 34.7 24.5 21.3 5.2 28.7 7.1 0.0 13.3 7.7 7.1 13.3 4.4 35.0 19.7 7.3 1.7 11.3 n.a. n.a. n.a. n.a. n.a. n.a. 7.1 13.3 5.9 35.0 1.9 10.4

1995–2000 0.9 0.7 5.1 5.9 16.6 18.0 14.3 3.6 2.5 9.6 211.1 3.0 2.4 18.4 6.7 6.1 5.4 4.1 13.6 10.3 2.0 9.6 5.7 9.0 19.1 8.7 7.5 0.6 4.5 13.1 5.3 3.1 3.7 12.7 n.a. 9.7

Note : Some of the growth ﬁgures might be misleading. For example, in the case of Equatorial Guinea the average growth rate of commercial services exports for 1995–2000 is estimated to be 211 per cent but still its average share in global commercial services exports declined during the same period. This is because the variation in Equatorial Guinea’s exports is very high. It exported 5, 6, 8, 9, 3 millions of services for the periods 1990–94 and 4, 5, 6, 1, 14, 10 for 1995–2000. Thus, average exports for 90–94 and 1995–2000 are 6.2 and 6.66 millions respectively or 1995–2000 average growth over 90–94 exports is 7.41 per cent. On the other hand the growth of world services exports of 1995–2000 over 90–94 was more than 47 per cent.

Marginalization and World Trade 750

Afghanistan

7500

7500

Angola

Bangladesh

600

150

Benin

500

5000

5000

400

100

250

2500

2500

200

50

1980

2000

1980

2000

1980 1000

300

Burkina Faso

150

200

100

100

50

Burundi

2000

1980 200

Cambodia

2000

Central Af.Rep.

150 500

Bhutan

1980 300

2000

Chad

200

100

100

50 1980 3000

2000

Congo, D.R.

2000

1980 1000

Ethiopia

2000

1980 600

Lao PDR

2000

300

200

1980

2000

Mauritania

400

1980 300

2000

Mozambique

Guinea-Bissau

1980

200

2000

Haiti

100

2000

2000

400

200

200

2000

Myanmar

Malawi

400

1980 1000

2000

1980

2000

1980

150

200

2000

100

Tanzania

1980

400

600

2000

Togo

1980

500

1980

Somalia 1000

2000

Sudan

100 500

50 2000

2000

Niger

200

50 1980

Mali

1980 600

Nepal

2000

400

Sierra Leone

Senegal

500

100

1500

1980 600

500

1980 1000

Rwanda

1980

2000

500

100 1980

1980 600

1000

200

200

2000

400

500

100

800

2000

25

1980

300

150

1980

50

Madagascar

Liberia

400

200

Guinea

500

200

600

2000

200 1980

300

600

2000

400

1000

400

1980

2000

1980

2000

1980

Yemen, Republic of 1500

Uganda

2000

Zambia

4000 1000

200

400

250

2000 500

1980

2000

1980

2000

1980

2000

1980

2000

1980

2000

Appendix 7.7. Merchandise Exports of Individual LDCs ($Million), 1970–2000

237

The Implications of Declining Commodity Prices 75

Antigua Barbuda 5000

50

2500

25 1980 20 15

Cape Verde

200 150

200

100

2000

Belize

2000

2000

1980

Djibouti

75

30

1000

2000

Dominica

50

20

500

Botswana

1000 1980

40

Cyprus

3000 2000

50 1980

1500

Comoros

10

5

25

10 1980

2000

1980 750

Eq Guinea

2000

1980

4000

Fiji

3000

500

2000

1980

75

Gabon

2000

1980

Gambia

30

250 1980

Guyana

2000 1500

2000

400

1980

2000

10 2000

Kiribati

1980 200

30

2000

1980

Lesotho

75

2000

Maldives

25

10 1980

2000

50

100

20

500

200

25 1980

40

Jamaica

1000

Grenada

20

1000 1980

2000

50

2000

200

600

1980

20

Barbados

300 100

2000

10

400

400

Bahrain

1980

2000

1980

2000

1980

2000

Sao Tome and Principe 2000

Malta

2000

20

PNGuinea

15

20

1000

10

10

2000

1980

150 100

50

50 1980

Suriname

2000

2000

100

Tonga

10

1980

2000

St Vincent & Grens.

50

1980

15

2000

1980

2000

1980

1980

4000 3000 2000 1000 2000

2000

Vanuatu 40 30 20

1980

2000

1980

Appendix 7.8. Merchandise Exports from Individual Small States (in $Million)

238

2000

50

5 1980

St Lucia

1980

Trinidad and Tobago Swaziland

500

500

2000

100

20 2000

1980 150

40

1980

1000

2000

St Kitts and Nevis

Solomon Islands Seychelles

30

Samoa

2000

5 1980

100

1000

3000

Mauritius

500 1980

150

1500 1000

1000

200

2000

2000

Marginalization and World Trade Afghanistan

250

15

200

10

150

5

100 1980

7.5

1990

2000

Burundi

1980

1990

2000

100

200

50 1980

20

Cambodia

1990

2000

Central African Rep.

2000

1990

2000

2000

1980

1990

2000

Congo, Dem. Rep. of 200

100 1990

2000

1980

50

Lao PDR

40

1990

2000

Liberia

30

50

5 1990

2000

Chad

1980

100 5

1980

20 1990

Guinea-Bissau

10

200

30

10 1980

10

Guinea

Burkina Faso

40

20

10 1990

40

50

Benin

1980

30 15

1980

15

Ethiopia

300

150

300

2000

50

2.5

400

1990

100

5

Bangladesh

400

1980

150

500

Angola

20 1980

1990

2000

1980

1990

2000

1980

1990

2000

1980

1990

2000

Mauritania 300

50

Madagascar

80

Malawi

40

30

Mali

300 200

20

30

40

100

100

20 1980

500

1990

2000

Myanmar

1980

750

1990

2000

1990

2000

200

40

1980

1990

2000

1980

400

1990

2000

Sudan

1990

2000

Yemen, Republic of Uganda

50 1990

2000

Senegal

300 200 1990

2000

1980

125

Tanzania

1990

2000

1990

2000

Togo

75 200 1990

2000

1980

1990

2000

1980

Z am bi a

100

100

100

1980

150

2000

400

1980

150

1990

100

200

1980

1980

400

Rwanda

1980

600

300

Somalia

100 2000

2000

20

20

1990

1990

10

25 1980

1980

20

40

50

2000

30

Sierra Leone 75

1990

30

250 1980

1980

Niger

Nepal

500 250

Mozambique

60

200

50 1980

1990

2000

1980

1990

2000

Appendix 7.9. Commercial Services Exports from Individual LDCs (in $Million)

239

The Implications of Declining Commodity Prices Antigua and Barbuda

400

1000

Bahrain 1000

300 200

150

Belize

750

100

Botswana

500

50

200

500

100

400

Barbados

300

750

100 1980

1990

2000

1980

1990

2000

1980

1990

2000

3000

100 Cape Verde

40

Comoros

1990

2000

2000

1980

1990

2000

100 Djibouti

Dominica

30 50

20

1980

1990

40 Cyprus

2000 50

1980

1000

1980

1990

2000

20

1980

1990

2000

1980

1990

2000

15

1980

1990

2000

150 Equatorial Guinea

10

600

Fiji

300

Gabon

100

1990

Grenada 100

400

200

200

100

50

5

1980

Gambia

2000

1980

1990

2000

1980

1990

2000

50

1980

1990

2000

1980

1990

2000

2000 Guyana

150

20

Jamaica 1500 1000

10

500

5

50 1980

1990

Malta

1000

2000

1980

1000

Kiribati

75

15

100

1990

2000

300

50

1980

1990

2000

400

Maldives

200

25

100 1980

1990

2000

1980

1990

2000

15

Papua New Guinea

Mauritius

Lesotho

50

Samoa

Sao Tome 10

500

500

1980

1990

2000

200

1980

75

1990

2000

1980

Solomon Islands

Seychelles

300

25

1990

2000

St Kitts and Nevis 100

50 50

25

2000

Suriname

1980

150

300

1990

2000

St Lucia

1990

2000

1980

1990

2000

15

Tonga

600

1990

2000

100

100

10

50

1990

2000

2000

Vanuatu

50

200 1980

1990

400

50 2000

2000

St Vincent and the Grenadines

1980

Trinidad and Tobago

100

1990

1990

50

1980

Swaziland

150

1980

1980

100

100

100 1990

1980

200

200

1980

5

1980

1990

2000

1980

1990

2000

1980

1990

2000

Appendix 7.10. Commercial Services Exports from Individual Small States (in $Million)

240

Marginalization and World Trade 1.5

Afghanistan

Angola

1

Bangladesh

Benin

1 1

.75

.5 2000

Burundi

.5

2000

Cambodia

1980

1

2000

Central African Rep.

1

1980

2000

2000

1

1

.5

2000

Guinea

Ethiopia 1.5

.5

1

1980

.5 1980

1

2000

Liberia

1980

1

2000

Guinea-Bissau

.5

2000

2000

2000

1980

2000

Senegal

1

1980

3

2000

2000

Sierra Leone

1 .5

Mauritania

2000

Rwanda

.5

1980

1

Somalia 1

2000

1980

1

Sudan

.5

.5

2000

1

1 1980

1980

Niger

2

.5 1980

2000

.5

.75

.75

2000

1980

1

Mali

1

.5

1980

Lao PDR

1

Nepal

Myanmar 1

.5

2000

.75

1980

Mozambique

2000

.5

1980

Malawi

.5 2000

1980

1

.5

1980

Madagascar

.5

1980

2000

Haiti

1 .5

1980

1

1980

1

2000

Congo, D.R.

.75

.5 1980

1980

Chad

.5

.5

1

1

.5 1980

1.5

1

1

.5

.5 1980

Burkina Faso

1

2000

Tanzania

.5

.5 1980

1.5

2000

1980

3

Togo

1

2

.5

1 1980

2000

2000

Uganda

1980

1.5

2000

1980

Yemen, Republic of

2000

1980

2000

Zambia 1

1 .5 .5 1980

2000

1980

2000

1980

2000

Appendix 7.11. Share of Individual LDCs in World Merchandise Exports, 1980 – 2000 Note: Share of individual LDCs in 1970 has been set to 1.

241

The Implications of Declining Commodity Prices 1.5

Antigua and Barbuda

Bahrain

Barbados

3

1.5

2

1

1

.5

1

2000

1980

Cape Verde

10

1

.5 1980

Belize

2000

1980

1

Comoros

5

.5 2000

Cyprus

1980

1

2000

Dominica

.5

.5

.5

.5 2000

1980

1

Eq Guinea 1 .5

2000

1 .75

1980

1980

Djibouti

1 .5

Botswana

1.5

1

2000

1980

Fiji 2.5

.75

2

.5

1.5

2000

1980

1

Gabon

2000

Gambia

1980

1

.5

2000

Grenada

.5

1 1980

1

2000

Guyana

1980

1

2000

1980

2000

Kiribati

Jamaica

1980

2

1.5 .5

1980

2000

1980

2000

2000

1.5

2.5

1

1.5

2000

Samoa

1

1980

1

2000

Sao Tome and Principe

1

2

2

.5

.5

1 1980

2000

1980

2

Seychelles

3

1.5

2

1

2000

1980

Solomon Islands

2000

St Kitts and Nevis.

1980

2000

Suriname

1.5 1

2000

Swaziland

.75

1980

1

2000

2000

2000

St Vincent & Grens.

2 1 1980

Tonga

2000

Trinidad and Tobago

1980

1

2000

Vanuatu

.5

1

.5

.5 1980

1980

3

St Lucia

.5

1.5 .5

2000

1

1980

1

1980

2

1

1

Maldives

.5

1980

PNGuinea

Mauritius

Malta 3

4

1980

2000

1

1

.5

1980

Lesotho

1.5

1

.5

2000

.5 1980

2000

1980

2000

1980

2000

1980

2000

Appendix 7.12. Share of Individual Small States in World Merchandise Exports, 1970–2000 Note: Share of individual small states in 1970 has been set to 1.

242

Marginalization and World Trade Afghanistan

Bangaladesh

Angola

5

1

1

.5

.5

1980

1990

2000

1980

1990

2000

2.5

Cambodia

1980

1990

2000

2000

.5

1980

1990

1990

2000

1980

2.5

Lao PDR

1990

2000

Liberia

1.5

2

2

1.5

1

1

1

.75

2000

1980

1990

2000

Mauritania

Mali

1

1990

2000

1

2000

Niger

2000

Malawi

1 .5

1990

2000

1980

2

1990

2000

1990

2000

Nepal

2 1

1980

1.5

1990

1.5

.5

.5 1990

1.5

Myanmar

.5

1980

Ethiopia

1980

3 .75

.75

2000

Madagascar

1980

Mozambique

1

1

2000

1 1990

1.25

1980

1990

1.5

1980

3

1980

2

Congo, D.R.

.5 1980

4

Guinea

2000

.5

1 1990

1990

1

1 1.5

1980

.7 1980

1.5

1.5

2 .5

.8

.5

Central African Rep.

Burundi 1

1.25

.9

.75

2.5

1

Burkina Foso

Benin 1

1990

Rwanda

2000

1 1980

1.25

1990

2000

Senegal

1980

1

1990

2000

Somalia

Sierra Leone 1

1 1

.5

.75

.5

1980

.5

.5 .5 1980

1.5

1990

2000

1980

1980

1.5

1

1

.75

.5

.5 1990

2000

Tanzania

Sudan

1980

1990

2000

1990

2000

1980

Togo 10

1990

2000

1980

1

Uganda

1990

2000

Zambia

1

1980

1990

2000

.5

5

.5

1980

1990

2000

1980

1990

2000

1980

1990

2000

Appendix 7.13. Share of Individual LDCs in World Commercial Services Exports, 1980 –2000 Note: Share of individual LDCs in 1980 has been set to 1.

243

The Implications of Declining Commodity Prices Antigua and Barbuda

Bahrain

3

2.5

Barbados

3

1.2 2

2

1

1 1980

1990

2000

1980

1990

1

1.5

.8

1

2000

1980

Cape Verde 3

5

2.5

Comoros

1990

2000

2.5

2000

1980

Eq Guinea 1.25

1

1990

2000

1990

2000

Guyana

2.5

1990

2000

1990

2000

1990

2000

Jamaica

1980

1990

1990

2000

Seychelles

1.2 1

2000

2 1 1980

1990

2000

1990

2000

4

1990

1990

2000

2000

1990

2000

Maldives

3 2

1990

2000

1990

2000

St Vincent and the Grenadines

1.5 1 1980

Tonga

1990

2000

1980

Trinidad and Tobago 1.2

1

2000

Sao Tome

1980

2

St Lucia

1 2000

1990

1 1980

1.5

1990

1980

Samoa

2

1

1990

2000

Vanuatu

1

.75

1980

1990

2000

.8

.5

.5 1990

2000

2

St Kitts and Nevis

1980

Swaziland

1

1980

Grenada

1980

1.5

2000

3

1.5 .5

2000

1 1990

1 1980

Solomon Islands

1980

Suriname

1

2000

1.5

1980

Papua New Guinea

2 2000

1990

Lesotho

3

1

1990

1990

.75

1980

4

.8 1980

1980

2

.5 1990

1

1.5

1980

Kiribati

1

1.5

2000

2000

1

Mauritius

.6

1 1990

1

1980

1.5

2 .8

Dominica

3

1

1980

Malta

2000

1.5

1 1980

4

Gambia

.5

1.2

1

1990

Djibouti

1980

Gabon

1.4

1.5

1980

1.5

1980

1.6

2

2000

2

.75 1980

1990

.5 1980

1

Fiji

1 .5

1980

.75

1 1990

Botswana

.75

1.5

1 1980

1.25 1

1

Cyprus

2

2

Belize

2

1980

1990

2000

1980

1990

2000

1980

1990

2000

Appendix 7.14. Share of Individual Small States in World Commercial Services Exports, 1980–2000 Note: Share of individual small states in 1980 has been set to 1.

244

Marginalization and World Trade Appendix 7.15. Absolute Changes in Merchandise Exports in the 1990s

World United States China Mexico Canada Korea, Rep. United Kingdom Russian Fed. France Taipei, Chinese Netherlands Japan Ireland Hong Kong, China Germany Saudi Arabia Malaysia Spain Philippines United Arab Emirates Iraq Singapore Norway Indonesia Hungary BelgiumLuxembourg Venezuela Thailand Israel Algeria India Iran, Islamic Rep. of Australia Austria Nigeria Vietnam Poland Ukraine Brazil Qatar Czech Rep. Sweden Kazakhstan Turkey Belarus Finland Kuwait Argentina

avg. exports 90–94

avg. change 90–94

avg. change 1995–2000

avg. change 1995–2000 as % of 90–94 avg. exports

3740196

200559

269113

7.20

448177 86338 48473 139956 78353 189361

29759 14729 5043 9437 7749 4962

39435 20103 17376 16888 9442 8577

8.80 23.28 35.85 12.07 12.05 4.53

27360 223025 80441 140555 340300 27894 117497

n/a 4665 6434 6542 27356 2584 17269

8436 7624 7351 7246 7227 6524 5762

30.83 3.42 9.14 5.16 2.12 23.39 4.90

412232 45501 42089 62843 10196 20696

1500 451 7357 4372 1309 258

5762 5509 4844 4728 4456 4260

1.40 12.11 11.51 7.52 43.71 20.58

2874 69220 33975 33132 10099 123835

2968 11019 161 3595 175 5437

4021 3907 3582 3341 3045 2854

139.92 5.64 10.54 10.09 30.15 2.30

15398 33201 13766 11148 20355 16250

352 5548 1201 1013 1762 520

2669 2524 2472 2358 2350 2309

17.33 7.60 17.95 21.15 11.54 14.21

42938 42380 11429 2819 14800 869 36196 3480 8515 56005 265 14944 193 25318 7260 13069

1947 912 1064 413 730 n/a 3036 169 n/a 938 n/a 1287 n/a 736 1055 827

2152 2036 1850 1800 1751 1726 1716 1589 1532 1510 1389 1228 1193 1159 1154 1066

5.01 4.80 16.19 63.85 11.83 198.62 4.74 45.65 17.99 2.70 524.15 8.21 618.13 4.58 15.90 8.16 (Continued )

245

The Implications of Declining Commodity Prices Appendix 7.15. (Continued ) avg. exports avg. change avg. change avg. change 1995–2000 as % 90–94 90–94 1995–2000 of 90–94 avg. exports Oman Libyan Arab Jamahiriya Angola Italy Slovak Rep. Bangladesh Colombia Morocco Lithuania Romania Costa Rica Chile Yemen South Africa Estonia Dominican Republic Trinidad and Tobago Portugal Turkmenistan Congo Sri Lanka Bahrain Yugoslavia Peru Azerbaijan El Salvador Sudan Egypt Latvia Brunei Darussalam Syrian Arab Republic Cambodia Uzbekistan Bosnia and Herzegovina Pakistan Equatorial Guinea Myanmar Denmark Gabon Ecuador Guatemala Macau, China Nepal Botswana Malta Slovenia Tunisia Kyrgyzstan ˆ te d’Ivoire Co Tajikistan Georgia Panama Cameroon Moldova, Rep. of Armenia

246

5330 10473 3574 175627 3701 1985 7290 4118 172 4980 2076 9624 692 24062 147 2823 1910 16872 96 1044 2509 3613 700 3622 75 849 388 3175 97 2338 3385 214 189 92 6731 70 532 39284 2235 3059 1304 1744 320 1807 1366 3918 3941 23 2798 76 8 467 1809 30 11

9 1231 217 5186 n/a 232 413 58 n/a 298 355 808 61 443 n/a 321 32 378 n/a 6 306 36 n/a 331 n/a 167 37 223 n/a 5 291 39 n/a n/a 487 0 118 1383 37 276 90 37 38 24 109 n/a 283 n/a 83 n/a n/a 61 129 n/a n/a

957 890 849 728 658 598 583 542 500 491 482 427 427 425 404 391 361 354 352 343 327 318 299 291 281 258 250 248 234 233 214 197 193 183 170 162 154 140 127 124 108 104 92 92 84 83 75 74 66 62 55 47 46 41 40

17.95 8.49 23.75 0.41 17.78 30.15 8.00 13.16 290.70 9.86 23.23 4.43 61.64 1.76 274.83 13.87 18.91 2.10 366.67 32.85 13.02 8.80 42.64 8.02 374.67 30.35 64.60 7.80 241.24 9.95 6.33 92.14 102.12 197.84 2.53 231.52 28.89 0.36 5.70 4.05 8.30 5.99 28.73 5.07 6.18 2.13 1.90 321.74 2.35 81.58 687.50 10.06 2.53 136.67 363.64

Marginalization and World Trade Mozambique Uruguay Bolivia Jordan Seychelles New Caledonia Iceland TFYR Macedonia Guyana Nicaragua Honduras Guinea Liberia French Polynesia Namibia Mali Albania Ethiopia Bhutan Haiti Belize Lesotho Guam Cuba Barbados Guinea-Bissau Senegal Maldives Grenada Aruba Cayman Islands Congo, Dem. Rep. of Lao PDR St Kitts and Nevis Kiribati Dominica St Vincent and the Grenadines Samoa Cook Islands Djibouti Paciﬁc Islands Cape Verde Uganda British Virgin Islands Montserrat Tuvalu Rwanda Chad Sao Tome and Principe Tonga Vanuatu Comoros St Pierre and Miquelon Bermuda Nauru Central African Republic Gambia Tanzania, United Rep. of Antigua and Barbuda

142 1712 849 1213 51 403 1538 668 336 289 816 675 334 144 1250 339 133 114 71 112 113 102 42 2377 193 19 717 45 23 43 16 608 169 26 4 52 67 6 4 18 18 5 218 3 2 0 75 170 6 13 22 19 28 53 43 131 46 428 38

6 55 27 90 1 24 8 n/a 51 5 3 11 5 29 59 6 23 n/a 2 20 5 21 0 929 7 4 8 2 1 3 1 145 56 2 1 2 8 1 0 2 0 0 69 0 0 0 17 8 1 1 2 2 8 7 7 8 1 26 2

39 38 26 26 26 24 23 23 23 21 20 20 20 20 17 15 12 12 12 11 10 10 8 7 7 6 6 5 5 5 4 3 2 2 2 2 1 1 1 1 0 0 0 0 0 0 0 0 0 1 1 1 1 1 2 2 2 4 4

27.64 2.21 3.06 2.11 50.20 5.96 1.52 3.44 6.85 7.28 2.50 2.96 5.99 13.63 1.34 4.54 9.02 10.51 16.43 9.63 9.19 9.78 18.10 0.29 3.42 32.98 0.78 11.95 23.08 12.09 24.05 0.43 1.47 8.46 50.00 3.08 1.49 16.13 23.81 3.26 2.22 7.69 0.09 6.67 0.00 0.00 0.27 0.12 7.14 4.55 2.75 4.12 3.55 1.87 3.72 1.22 3.90 0.89 10.53 (Continued )

247

The Implications of Declining Commodity Prices Appendix 7.15. (Continued )

Togo Afghanistan Niger Sierra Leone Bahamas Korea, DPR Somalia Suriname Benin Paraguay Mauritius American Samoa Ghana Burundi St Lucia Solomon Islands Malawi Burkina Faso Fiji Greenland Lebanon Madagascar Switzerland Jamaica Kenya Swaziland Mongolia Mauritania Greece Zimbabwe Croatia Netherlands Antilles Zambia Cyprus New Zealand Bulgaria Papua New Guinea

248

avg. exports 90–94

avg. change 90–94

avg. change 1995–2000

avg. change 1995–2000 as % of 90–94 avg. exports

219 173 288 133 1363 1265 122 399 347 779 1266 319 1033 82 117 105 384 172 488 349 503 314 64905 1114 1285 658 426 423 8823 1631 2554 1521 980 949 10325 3985 1885

27 20 14 6 5 254 4 6 28 36 38 2 132 8 5 18 22 23 16 36 6 22 1644 19 145 57 76 21 324 40 n/a 103 96 3 674 256 374

4 5 5 6 6 6 7 8 9 10 10 10 11 11 13 15 17 19 19 20 20 21 21 26 29 29 34 34 35 39 40 46 50 55 76 106 110

1.83 2.90 1.81 4.36 0.44 0.51 5.72 2.10 2.54 1.23 0.76 3.13 1.05 13.35 10.92 14.26 4.53 10.81 3.98 5.78 4.02 6.82 0.03 2.35 2.23 4.44 7.97 8.12 0.40 2.37 1.57 3.05 5.06 5.80 0.73 2.66 5.82

Marginalization and World Trade Appendix 7.16. Absolute Changes in Commercial Services Exports in the 1990s

World United States United Kingdom Spain Ireland China India Canada Greece BelgiumLuxembourg Hong Kong, China Korea, Rep. Israel Netherlands Germany Taipei, Chinese Denmark Turkey Sweden Japan Mexico Brazil Australia Malaysia Norway Cuba Croatia Egypt Saudi Arabia Dominican Republic United Arab Emirates Switzerland Hungary Ukraine Morocco Iran, Islamic Rep. of Argentina Bulgaria Costa Rica Kuwait Estonia Chile Lithuania Vietnam South Africa Belarus

avg. exports 90–94

avg. change 90–94

avg. change 1995–2000

avg. change 95–99 as % of avg. 90–94 exports

903144 157568.0 59235.0 30708.8 3669.4 9813.2 5094.4 20470.2 7918.4 28978.2

64086 12270.0 3429.3 1454.0 214.8 2651.5 355.5 1215.0 657.0 2675.5

53298 14700.0 7777.0 2656.2 2367.8 2343.2 2181.4 2172.4 1930.6 1777.8

5.90 9.33 13.13 8.65 64.53 23.88 42.82 10.61 24.38 6.13

24364.8 11607.2 5492.4 35099.2 55879.4 10362.0 13411.6 9309.6 13878.6 48171.4 8323.8 3940.8 11341.8 5685.6 12339.8 801.6 1388.0 6555.2 3207.0 1367.6

3253.5 1769.5 500.8 3002.0 1851.5 1544.5 211.8 710.3 17.5 3848.0 713.3 277.8 983.5 1357.8 134.3 134.0 714.3 720.0 79.0 162.8

1551.4 1355.4 1309.4 1191.6 1059.6 1058.6 1053.4 951.4 935.6 867.4 796.4 568.2 430.4 428.2 367.2 344.4 325.8 285.0 261.0 249.8

6.37 11.68 23.84 3.39 1.90 10.22 7.85 10.22 6.74 1.80 9.57 14.42 3.79 7.53 2.98 42.96 23.47 4.35 8.14 18.27

2480.8

128.0

241.6

9.74

20058.4 2811.6 2450.2 1813.0 548.4

916.5 92.3 133.5 1.5 16.8

232.2 216.4 190.8 166.6 164.8

1.16 7.70 7.79 9.19 30.05

2688.0 947.0 844.2 1059.0 279.6 2266.8 216.4 682.0 3245.2 189.4

233.8 105.0 145.0 33.8 82.3 244.5 35.8 275.3 66.3 22.8

156.6 139.6 139.0 130.0 125.4 118.8 114.0 111.0 103.4 103.2

5.83 14.74 16.47 12.28 44.85 5.24 52.68 16.28 3.19 54.49 (Continued )

249

The Implications of Declining Commodity Prices Appendix 7.16. (Continued )

Panama Latvia Bahamas Qatar Kazakhstan Peru Jamaica Aruba Iceland Ghana Colombia Albania Nigeria Algeria El Salvador Trinidad and Tobago Mauritius Oman Romania Zimbabwe Cyprus Barbados Tunisia Honduras Turkmenistan Myanmar Portugal Bahrain Nicaragua Cameroon TFYR Macedonia Botswana Sri Lanka Maldives Armenia Guatemala Ecuador Haiti Seychelles Madagascar Mozambique Uganda Gambia Ethiopia Namibia Congo Azerbaijan Malta Antigua and Barbuda Cambodia

250

avg. 90–94

avg. change 90–94

avg. change 1995–2000

avg. change 1995–2000 as % of avg. 90–94 exports

1039.6 401.0 1412.6 354.4 424.4 780.2 1129.8 522.8 475.0 110.4 1799.8 42.4 887.6 529.4 310.4 364.6

66.3 96.8 6.3 15.8 22.3 59.0 119.0 51.8 21.8 14.3 6.5 11.8 148.5 14.3 9.3 1.3

101.6 95.0 94.4 94.4 90.4 84.2 84.0 77.8 71.2 70.2 69.8 67.0 66.6 62.0 61.4 59.4

9.77 23.69 6.68 26.64 21.30 10.79 7.43 14.88 14.99 63.59 3.88 158.02 7.50 11.71 19.78 16.29

552.6 33.6 754.2 297.4 1997.6 652.0 1779.4 172.8 61.2 147.8 5830.6 556.8 60.0 364.8 72.2 177.2 579.8 142.8 11.0 519.8 571.8 31.4 177.4 146.6 157.2 42.8 65.4 247.0 167.2 54.6 130.4 846.2 342.6

37.8 13.8 103.3 25.0 159.5 39.5 149.8 21.5 3.0 40.8 411.8 115.0 14.8 15.0 32.0 2.0 76.0 23.5 0.5 86.5 33.0 9.0 6.0 13.5 22.0 16.0 5.8 1.3 34.8 4.0 7.3 63.5 20.8

58.8 54.0 52.8 49.8 43.0 42.2 40.2 38.2 32.6 31.8 31.2 29.4 29.2 28.2 26.8 23.4 23.0 23.0 22.8 21.4 21.2 21.2 21.0 19.0 16.6 16.4 15.6 15.4 14.6 14.6 13.6 12.2 11.6

10.64 160.71 7.00 16.75 2.15 6.47 2.26 22.11 53.27 21.52 0.54 5.28 48.67 7.73 37.12 13.21 3.97 16.11 207.27 4.12 3.71 67.52 11.84 12.96 10.56 38.32 23.85 6.23 8.73 26.74 10.43 1.44 3.39

48.0

0.8

11.2

23.33

Marginalization and World Trade St Vincent and the Grenadines Tanzania, United Rep. of Grenada St Lucia Lao People’s Dem. Rep. Vanuatu Cape Verde Angola Guyana Belize Mongolia Fiji Bolivia Gabon Rwanda Dominica Libyan Arab Jamahiriya Solomon Islands Yemen Guinea Comoros Kyrgyzstan Uruguay Sierra Leone Moldova, Rep. of Uzbekistan St Kitts and Nevis Malawi Mauritania Samoa Lesotho Equatorial Guinea Sao Tome and Principe Congo, Dem. Rep. Liberia Burkina Faso Guinea-Bissau Zambia Tajikistan Paraguay Kiribati Tonga Niger Afghanistan Yugoslavia Czech Rep. Czech and Slovak Fed. Rep., former Djibouti Mali Somalia Burundi

52.6

5.0

10.4

19.77

241.2

70.0

9.8

4.06

79.2 190.2 35.4

9.3 22.0 11.3

9.4 8.6 8.6

11.87 4.52 24.29

63.0 32.4 103.2 100.8 101.0 36.6 419.8 153.6 260.6 24.4 41.6 55.8

3.5 2.3 21.3 8.0 6.5 3.3 28.0 11.0 3.5 2.3 4.5 14.5

8.4 8.4 7.4 7.2 7.0 6.8 6.6 6.6 6.0 5.6 5.0 4.6

13.33 25.93 7.17 7.14 6.93 18.58 1.57 4.30 2.30 22.95 12.02 8.24

30.0 106.0 64.4 15.2 12.8 843.2 52.6 28.8 246.4 75.2 31.4 15.0 33.4 32.0 6.2

5.8 5.5 19.3 3.8 6.5 216.0 10.3 2.0 17.3 9.5 3.8 0.8 1.5 1.0 0.5

4.4 4.0 3.8 3.8 3.6 3.4 3.0 3.0 2.6 2.6 2.4 2.0 1.6 1.2 1.2

14.67 3.77 5.90 25.00 28.13 0.40 5.70 10.42 1.06 3.46 7.64 13.33 4.79 3.75 19.35

4.4

0.5

1.2

27.27

145.2

42.3

1.0

0.69

41.0 38.4 6.4 75.0 41.4 401.6 13.0 12.6 17.4 4.8 1767.8 1959.8 1864.2

2.5 1.0 0.5 4.5 2.8 1.8 2.3 0.8 3.5 1.0 1593.5 1280.0 640.8

0.8 0.8 0.8 0.6 0.6 0.4 0.4 0.4 0.2 0.2 0.0 0.0 0.0

1.95 2.08 12.50 0.80 1.45 0.10 3.08 3.17 1.15 4.17 0.00 0.00 0.00

33.6 60.8 0.0 7.0

0.5 5.0 0.0 0.3

0.0 0.0 0.0 0.4

0.00 0.00 0.00 5.71 (Continued )

251

The Implications of Declining Commodity Prices Appendix 7.16. (Continued ) avg. change 95–99 as avg. 90–94 avg. change 90–94 avg. change 1995–2000 % of avg. 90–94 exports Central African Republic Benin Chad Montserrat Togo Senegal Suriname Papua New Guinea Georgia ˆ te d’Ivoire Co Sudan Jordan Swaziland Bosnia and Herzegovina New Zealand Slovenia Pakistan Kenya Syrian Arab Republic Slovak Rep. Netherlands Antilles Nepal Bangladesh Poland Indonesia Brunei Darussalam Venezuela Macau, China Thailand Russian Fed. Finland Austria France Singapore Philippines Italy

252

16.6

0.0

0.6

3.61

117.4 18.8 21.2 80.2 313.0 43.0 267.2

1.3 0.0 2.3 15.5 11.8 9.8 9.3

0.8 1.6 1.8 2.4 3.0 3.2 8.2

0.68 8.51 8.49 2.99 0.96 7.44 3.07

203.6 445.2 77.8 1460.2 95.6 306.8

12.3 1.5 22.5 28.3 2.0 132.5

11.0 11.4 11.6 13.2 15.6 21.4

5.40 2.56 14.91 0.90 16.32 6.98

2781.4 883.4 1310.4 767.4 1138.0

296.0 451.8 56.5 10.3 219.0

26.0 27.0 29.6 30.0 30.2

0.93 3.06 2.26 3.91 2.65

832.0 1265.6

555.3 63.0

32.0 34.2

3.85 2.70

277.6 377.4 4503.2 3451.8 409.2

90.3 30.8 863.8 548.0 40.8

36.4 37.2 49.4 56.4 77.6

13.11 9.86 1.10 1.63 18.96

1231.8 2137.8 8977.6 7460.0 4577.8 25717.0 71874.8 16797.4 4428.2 51414.4

83.3 312.5 1283.3 447.8 213.3 1184.8 2103.5 2555.0 963.0 1147.8

92.4 113.0 173.4 187.2 266.4 329.8 391.0 519.2 1038.0 1123.0

7.50 5.29 1.93 2.51 5.82 1.28 0.54 3.09 23.44 2.18

Marginalization and World Trade Appendix 7.17. Absolute Changes in Total Exports (Merchandise plus Commercial Services) in the 1990s Countries World United States China Canada Mexico United Kingdom Korea, Rep. Taipei, Chinese Ireland Hong Kong, China Japan Spain Germany Saudi Arabia Malaysia India Russian Fed. BelgiumLuxembourg United Arab Emirates Norway Netherlands Singapore Israel Indonesia France Philippines Turkey Greece Hungary Venezuela Australia Sweden Thailand Brazil Iran, Islamic Rep. of Poland Vietnam Czech Rep. Kuwait Qatar Argentina Denmark Oman Austria Costa Rica Kazakhstan Algeria

avg. exports avg. change avg. exports avg. change avg. change 1995–2000 as 90–94 90–94 1995–2000 1995–2000 % of avg. 90–94 4637277 599514 81458 162231 56796 248470

269602 41303 15412 10381 5756 9525

6883661 902478 205487 259261 129147 366896

297624 54262 26547 20387 18179 16234

6.42 9.05 32.59 12.57 32.01 6.53

89098 90704 31165 141732 380450 94228 467619 48707 46781 25882 67284 141801

9596 7933 2790 20566 29696 6021 6736 371 8381 2165 3731 5827

166030 138230 72391 220417 477397 154249 611378 60678 93206 46729 97934 193571

11619 8329 8170 7339 7027 6985 6176 5770 5726 4560 4343 4130

13.04 9.18 26.21 5.18 1.85 7.41 1.32 11.85 12.24 17.62 6.46 2.91

24808

1279

35990

4101

16.53

46595 168655 88468 19645 36865 286955 14770 24427 13897 11752 16805 54307 69136 41781 40451 19618

42 5954 13365 1626 3902 7573 2287 2051 377 284 296 2916 824 6700 3451 49

61882 246131 149159 33523 58898 372970 36330 47401 19207 24845 24293 76194 102476 72061 56755 21272

3917 3698 3576 3544 3534 2990 2932 2889 2787 2717 2698 2595 2441 2327 2284 2162

8.41 2.19 4.04 18.04 9.59 1.04 19.85 11.83 20.06 23.12 16.06 4.78 3.53 5.57 5.65 11.02

19722 3339 7999 8292 3543 15893 51821 5402 67863 2560 5805 11740

1493 856 5271 1108 157 1151 1629 5 2375 337 227 933

40087 11953 31577 15338 6025 28812 65942 7754 91861 6557 7497 13255

2123 1961 1508 1479 1287 1206 1124 1105 1068 1055 925 773

10.76 58.72 18.86 17.83 36.32 7.59 2.17 20.45 1.57 41.20 15.94 6.58 (Continued )

253

The Implications of Declining Commodity Prices Appendix 7.17. (Continued ) Countries Egypt Finland Colombia Dominican Republic Slovak Rep. Nigeria Chile Romania Belarus Bangladesh Ukraine Estonia Angola Cuba Lithuania South Africa Peru Sri Lanka Bahrain El Salvador Aruba Croatia Yemen Congo Morocco Azerbaijan Equatorial Guinea Sudan Latvia Syrian Arab Republic Guatemala Bahamas Botswana Portugal Honduras Myanmar Ghana Ecuador Trinidad and Tobago Zimbabwe Malta Tunisia Cambodia Cameroon Gabon Iceland Madagascar Bosnia and Herzegovina Haiti Albania

254

avg. exports avg. change avg. exports avg. change avg. change 1995–2000 as 90–94 90–94 1995–2000 1995–2000 % of avg. 90–94 10425 30017 9467 3092

750 1051 489 842

14151 48928 13718 7163

763 763 675 641

7.32 2.54 7.13 20.74

3264 12287 11892 5843 2212 2565 14705 949 3520 3334 2186 27986 4481 3010 4173 1133 1434 3807 1429 1205 6775 583 56

2232 1180 1052 199 201 346 806 284 196 887 81 566 378 415 79 161 337 1779 109 112 330 32 6

12140 14013 19852 10066 6867 5223 18665 3662 4856 3986 4498 35022 7607 5419 5100 2799 2502 8080 2485 1734 9623 1175 544

629 568 546 544 529 496 486 445 403 397 383 376 372 351 347 326 325 313 312 303 276 263 240

19.27 4.62 4.59 9.31 23.91 19.34 3.31 46.86 11.46 11.92 17.51 1.34 8.30 11.67 8.31 28.74 22.66 8.21 21.86 25.12 4.08 45.09 430.47

412 1271 4593

27 171 12

870 2806 5504

239 233 227

57.86 18.34 4.95

1847 1637 1978 22985 1108 646 1147 3721 2105

171 15 18 959 73 199 100 336 47

3230 2030 2732 33416 2126 1616 2049 5485 2979

206 187 182 176 171 169 164 155 146

11.16 11.45 9.21 0.77 15.39 26.17 14.26 4.17 6.94

2005 2279 5708 353 2160 2594 2001 500 589

78 169 432 63 183 35 15 47 254

3026 3052 8235 1037 2306 2992 2661 844 1287

126 118 114 106 105 97 91 82 80

6.30 5.16 2.00 29.88 4.84 3.74 4.53 16.49 13.65

161 186

61 34

386 397

80 77

49.94 41.51

Marginalization and World Trade Brunei Darussalam TFYR Macedonia Nicaragua Slovenia Mauritius Mozambique Guyana Panama Nepal Barbados Uruguay Pakistan Mali Namibia Bolivia Seychelles Guinea Jamaica Bulgaria Netherlands Antilles Afghanistan Armenia Fiji Ethiopia Maldives Lao People’s Dem. Rep. Senegal Mongolia Georgia Kyrgyzstan Jordan ˆ te d’Ivoire Co Grenada Tajikistan Gambia Belize Uganda Lesotho Solomon Islands Antigua and Barbuda Togo Vanuatu Benin St Vincent and the Grenadines Rwanda Guinea-Bissau Tanzania, United Rep. of St Kitts and Nevis Cape Verde Uzbekistan

2747

40

3021

70

2.56

584

263

1386

70

11.95

346 4803 1860 299 496 5860 605 851 2593 7838 416 1405 934 229 655 2361 5241 1584

17 2160 73 28 35 741 128 32 272 485 5 93 50 5 58 208 440 75

748 10604 2571 533 727 7580 1066 1230 3812 9625 607 1746 1331 366 710 3396 6326 1947

66 65 56 56 54 51 51 50 50 47 45 45 44 43 37 36 36 35

18.97 1.35 3.02 18.60 10.90 0.87 8.40 5.92 1.92 0.60 10.81 3.19 4.76 18.62 5.68 1.52 0.68 2.18

361 174 818 487 212 208

45 19 47 21 23 68

524 362 1096 836 399 431

32 31 30 28 28 27

8.86 18.00 3.69 5.75 13.10 12.79

1154 413 412 343 2677 3241

42 23 24 14 118 6

1349 545 591 561 3561 4792

26 24 24 24 23 22

2.22 5.91 5.93 6.99 0.85 0.68

105 350 202 238 276 134 135 397

8 101 9 13 87 20 24 24

167 797 217 320 719 230 216 437

21 20 20 16 15 11 10 9

20.46 5.61 9.70 6.90 5.44 8.48 7.25 2.37

488 81 467 120

62 6 42 4

483 126 582 156

9 9 8 8

1.80 10.89 1.80 7.00

98 25 670

20 3 98

96 47 1229

8 8 6

8.16 30.16 0.96

105

10

138

6

5.88

46 2962

0 167

99 3910

6 5

12.07 0.18 (Continued )

255

The Implications of Declining Commodity Prices Appendix 7.17. (Continued ) Countries Dominica Sierra Leone Comoros Samoa Congo, Dem. Rep. Zambia Tonga Central African Republic Liberia Burkina Faso Kiribati Sao Tome and Principe Niger Czech and Slovak Fed. Rep., former Djibouti Somalia Chad St Lucia Montserrat Mauritania Suriname Malawi Cyprus Burundi Swaziland New Zealand Moldova, Rep. of Kenya Papua New Guinea Turkmenistan Macau, China Italy Libyan Arab Jamahiriya Paraguay Switzerland

256

avg. exports avg. change avg. exports avg. change avg. change 1995–2000 as 90–94 90–94 1995–2000 1995–2000 % of avg. 90–94 94 189 34 40 1669 1276 27 151

3 2 2 0 310 17 3 1

133 115 44 71 1634 1469 36 163

5 3 3 3 2 2 2 1

5.30 1.59 8.72 6.57 0.14 0.14 6.67 0.93

530 284 18 8

17 15 3 1

618 268 26 13

1 1 1 1

0.23 0.35 4.55 9.52

360 8603

69 3550

280 0

0 0

0.11 0.00

89 66 197 310 23 433 573 424 2943 87 747 13006 574 2004 2239

3 2 24 14 3 10 125 15 164 2 62 1043 32 122 378

58 77 270 369 28 454 469 506 3873 71 1000 17636 815 2670 2557

0 0 2 3 4 5 7 10 13 13 27 40 49 60 85

0.00 0.00 0.91 1.10 17.39 1.20 1.15 2.27 0.43 15.24 3.61 0.31 8.61 3.00 3.80

1255 3895 227114 9860

69 354 6427 761

1279 5036 302716 8208

100 108 175 308

7.99 2.78 0.08 3.13

2863 98100

318 2065

3908 119927

369 537

12.88 0.55

Marginalization and World Trade Appendix 7.18. Average Merchandise Export Share of Individual Countries Average Share Millions of US Dollars

per cent of world exports

avg. 1985–89 avg. 1990–94 avg. 1995–2000 1985–89 World Canada United States Greenland Bermuda St Pierre and Miquelon Antigua and Barbuda Argentina Aruba Bahamas Barbados Belize Bolivia Brazil British Virgin Islands Cayman Islands Chile Colombia Costa Rica Cuba Dominica Dominican Republic Ecuador El Salvador Grenada Guatemala Guyana Haiti Honduras Jamaica Mexico Montserrat Netherlands Antilles Nicaragua Panama Paraguay Peru St Kitts and Nevis St Lucia St Vincent and the Grenadines Suriname Trinidad and Tobago Uruguay Venezuela Austria Belgium

2511416 103677 277267 317 40 26 18 8064 21 777 228 92 651 28418 3 11 5670 4810 1183 5711 44 1648 2310 626 29 1044 224 178 825 732 28405 3 1300 273 333 482 2872 26 89 68

3753796 139956 448177 349 53 28 38 13069 43 1363 193 113 849 36196 3 16 9624 7290 2076 2377 52 2823 3059 849 23 1304 336 112 816 1114 48473 2 1521 289 467 779 3622 26 117 67

5563180 222943 677504 304 55 6 16 24537 41 962 265 162 1132 50243 5 26 16116 11284 4901 1678 53 4718 4675 2293 32 2368 532 143 1355 1340 117708 3 1350 608 740 946 6202 26 76 47

384 1600 1238 11446 25970 -

399 1910 1712 15398 42380 -

489 2825 2422 22054 61717 60736

1990–94 1995–2000

100.00000 100.00000 100.00000 4.12822 3.72839 4.00748 11.04026 11.93930 12.17837 0.01263 0.00930 0.00547 0.00158 0.00142 0.00098 0.00104 0.00075 0.00010 0.00071 0.00101 0.00028 0.32111 0.34814 0.44105 0.00083 0.00115 0.00074 0.03095 0.03632 0.01729 0.00908 0.00515 0.00477 0.00366 0.00302 0.00291 0.02591 0.02262 0.02034 1.13155 0.96426 0.90314 0.00011 0.00008 0.00010 0.00042 0.00042 0.00046 0.22578 0.25639 0.28969 0.19153 0.19420 0.20283 0.04711 0.05531 0.08809 0.22739 0.06332 0.03016 0.00174 0.00139 0.00096 0.06564 0.07520 0.08481 0.09199 0.08150 0.08404 0.02494 0.02263 0.04122 0.00117 0.00062 0.00057 0.04155 0.03474 0.04256 0.00891 0.00895 0.00957 0.00708 0.00299 0.00256 0.03286 0.02175 0.02435 0.02914 0.02968 0.02409 1.13102 1.29130 2.11584 0.00010 0.00006 0.00006 0.05175 0.04052 0.02426 0.01088 0.00769 0.01093 0.01327 0.01245 0.01329 0.01918 0.02075 0.01701 0.11436 0.09649 0.11148 0.00104 0.00069 0.00046 0.00356 0.00312 0.00136 0.00270 0.00178 0.00084 0.01529 0.06370 0.04929 0.45575 1.03406 0.00000

0.01064 0.05088 0.04560 0.41021 1.12898 0.00000

0.00878 0.05077 0.04353 0.39643 1.10939 1.09176 (Continued )

257

The Implications of Declining Commodity Prices Appendix 7.18. (Continued ) Average Share Millions of US Dollars

per cent of world exports

avg. 1985–89 avg. 1990–94 avg. 1995–2000 1985–89 1990–94 1995–2000 BelgiumLuxembourg Bosnia and Herzegovina Croatia Denmark Finland France Germany Greece Iceland Ireland Malta Netherlands Norway Portugal Slovenia Spain Sweden Switzerland Turkey United Kingdom Albania Armenia Azerbaijan Belarus Bulgaria Czech Rep. Estonia Georgia Hungary Kazakhstan Kyrgyzstan Latvia Lithuania Moldova, Rep. of Poland Romania Russian Fed. Slovak Rep. Tajikistan Turkmenistan Ukraine USSR, former Uzbekistan Algeria Angola Benin Botswana Burkina Faso

258

79560

123835

114727

92

596

— 24427 19011 144438 277236 5939 1219 15681 612 90631 21813 9207 — 34096 42705 42526 9778 127422

2554 39284 25318 223025 412232 8823 1538 27894 1366 140555 33975 16872 3918 62843 56005 64905 14944 189361

— — — — 15362 — — — 9378 — — — — — 12640 10461 — — — — — 102192 — 9256 2255 218 1300 109

133 11 75 193 3985 8515 147 8 10099 265 23 97 172 30 14800 4980 27360 3701 76 96 869 21640 189 11148 3574 347 1807 172

—

3.16795 3.29892

2.06225

—

0.00246

0.01071

4432 50002 42171 303474 532984 11055 1921 59930 1904 207326 47533 24135 8564 104450 84052 79887 25410 269215

— 0.97263 0.75699 5.75124 11.03905 0.23647 0.04853 0.62439 0.02437 3.60877 0.86856 0.36661 0.00000 1.35762 1.70043 1.69332 0.38936 5.07373

0.06804 1.04651 0.67447 5.94133 10.98174 0.23505 0.04098 0.74309 0.03640 3.74435 0.90509 0.44948 0.10439 1.67413 1.49195 1.72904 0.39809 5.04452

0.07966 0.89880 0.75804 5.45505 9.58057 0.19871 0.03453 1.07727 0.03422 3.72675 0.85443 0.43383 0.15394 1.87753 1.51086 1.43600 0.45675 4.83923

214 230 838 5789 4714 24492 2501 209 20630 5794 453 1535 3175 582 26728 8596 82725 9979 675 1378 12228 0 3240 13525 5055 431 2452 289

0.00000 0.00000 0.00000 0.00000 0.61169 0.00000 0.00000 0.00000 0.37341 0.00000 0.00000 0.00000 0.00000 0.00000 0.50330 0.41654 0.00000 0.00000 0.00000 0.00000 0.00000 4.06910 0.00000 0.36854 0.08980 0.00869 0.05176 0.00435

0.00354 0.00029 0.00200 0.00514 0.10616 0.22684 0.00392 0.00021 0.26903 0.00706 0.00061 0.00258 0.00458 0.00080 0.39427 0.13267 0.72886 0.09859 0.00202 0.00256 0.02315 0.57648 0.00503 0.29698 0.09520 0.00924 0.04813 0.00458

0.00385 0.00413 0.01507 0.10406 0.08474 0.44025 0.04495 0.00376 0.37083 0.10415 0.00815 0.02759 0.05707 0.01047 0.48045 0.15451 1.48701 0.17938 0.01213 0.02476 0.21979 0.00000 0.05824 0.24311 0.09087 0.00774 0.04408 0.00519

Marginalization and World Trade Burundi Cameroon Cape Verde Central African Republic Chad Comoros Congo Congo, Dem. Rep. of ˆ te d’Ivoire Co Djibouti Egypt Equatorial Guinea Ethiopia Ethiopia, former Gabon Gambia Ghana Guinea Guinea-Bissau Kenya Lesotho Liberia Libyan Arab Jamahiriya Madagascar Malawi Mali Mauritania Mauritius Morocco Mozambique Namibia Niger Nigeria Rwanda Sao Tome and Principe Senegal Seychelles Sierra Leone Somalia South Africa Sudan Swaziland Tanzania, United Rep. of Togo Tunisia Uganda Zambia Zimbabwe Bahrain Cyprus Iran, Islamic Rep. of Iraq Israel Jordan Kuwait Lebanon Oman

114 984 6 110 114 17 900 1080 3047 22 2155 37 0 407 1460 41 897 534 14 1032 45 416 8750 302 266 194 391 792 2871 92 975 284 8389 126 6 615 27 130 95 19992

82 1809 5 131 170 19 1044 608 2798 18 3175 70 114 131 2235 46 1033 675 19 1285 102 334 10473 314 384 339 423 1266 4118 142 1250 288 11429 75 6 717 51 133 122 24062

67 1732 12 167 218 8 1703 516 4247 22 3715 488 486 0 2802 16 1719 749 40 1915 187 480 9427 279 460 514 417 1604 6763 241 1335 287 14609 63 5 975 119 22 131 28527

478 370 307

388 658 428

825 914 665

225 2192 337 940 1405 2554 598 10236 11872 8538 916 9054 553 3758

219 3941 218 980 1631 3613 949 16250 2874 13766 1213 7260 503 5330

346 5670 514 883 2165 4385 1146 19535 7663 23725 1825 14038 756 7439

0.00452 0.03917 0.00022 0.00440 0.00453 0.00069 0.03582 0.04301 0.12133 0.00088 0.08580 0.00146 0.00000 0.01622 0.05815 0.00162 0.03571 0.02128 0.00055 0.04109 0.00180 0.01655 0.34840 0.01203 0.01058 0.00774 0.01555 0.03155 0.11432 0.00367 0.03881 0.01132 0.33402 0.00502 0.00023 0.02450 0.00107 0.00516 0.00377 0.79604

0.00220 0.04819 0.00014 0.00348 0.00454 0.00052 0.02781 0.01620 0.07453 0.00049 0.08459 0.00186 0.00304 0.00350 0.05953 0.00123 0.02752 0.01798 0.00050 0.03422 0.00272 0.00890 0.27899 0.00835 0.01023 0.00904 0.01128 0.03374 0.10969 0.00378 0.03329 0.00767 0.30445 0.00200 0.00015 0.01909 0.00136 0.00354 0.00326 0.64101

0.00120 0.03113 0.00021 0.00300 0.00392 0.00014 0.03061 0.00928 0.07634 0.00039 0.06677 0.00878 0.00874 0.00000 0.05037 0.00028 0.03090 0.01346 0.00072 0.03442 0.00335 0.00863 0.16946 0.00501 0.00827 0.00925 0.00750 0.02882 0.12157 0.00434 0.02400 0.00515 0.26259 0.00112 0.00008 0.01753 0.00213 0.00040 0.00235 0.51278

0.01903 0.01033 0.01475 0.01752 0.01224 0.01141

0.01484 0.01643 0.01195

0.00896 0.08729 0.01343 0.03743 0.05594 0.10169 0.02382 0.40758 0.47273 0.33998 0.03646 0.36052 0.02204 0.14964

0.00621 0.10193 0.00924 0.01588 0.03891 0.07881 0.02061 0.35115 0.13775 0.42646 0.03281 0.25233 0.01358 0.13372

0.00583 0.10498 0.00582 0.02611 0.04344 0.09625 0.02527 0.43290 0.07656 0.36672 0.03231 0.19339 0.01340 0.14199

(Continued )

259

The Implications of Declining Commodity Prices Appendix 7.18. (Continued ) Average Share Millions of US Dollars

per cent of world exports

avg. 1985–89 avg. 1990–94 avg. 1995–2000 1985–89 1990–94 1995–2000 Qatar Saudi Arabia Syrian Arab Republic United Arab Emirates Yemen Afghanistan American Samoa Australia Bangladesh Bhutan Brunei Darussalam Cambodia China Cook Islands Fiji French Polynesia Guam Hong Kong, China India Indonesia Japan Kiribati Korea, Dem. People’s Rep. of Korea, Rep. of Lao PDR Macau, China Malaysia Maldives Mongolia Myanmar Nauru Nepal New Caledonia New Zealand Paciﬁc Islands Pakistan Papua New Guinea Philippines Samoa Singapore Solomon Islands Sri Lanka Taipei, Chinese Thailand Tonga Tuvalu Vanuatu Vietnam

260

2430 24725 1726 13370 526 452 284 28430 1124 51 2052 28 39594 4 390 66 62 50340 11802 18685 231599 4 1544

3480 45501 3385 20696 692 173 319 42938 1985 71 2338 214 86338 4 488 144 42 117497 20355 33132 340300 4 1265

6251 56445 3744 30398 2523 142 315 58694 4646 113 2563 604 185124 6 598 254 49 182489 35038 51385 426920 8 708

0.09675 0.98450 0.06873 0.53237 0.02094 0.01801 0.01130 1.13205 0.04477 0.00204 0.08172 0.00111 1.57654 0.00018 0.01551 0.00261 0.00248 2.00444 0.46992 0.74400 9.22185 0.00014 0.06148

0.09270 1.21212 0.09019 0.55134 0.01845 0.00460 0.00850 1.14386 0.05287 0.00188 0.06228 0.00570 2.30002 0.00011 0.01299 0.00383 0.00112 3.13009 0.54226 0.88263 9.06550 0.00010 0.03371

0.11236 1.01461 0.06731 0.54642 0.04535 0.00255 0.00566 1.05504 0.08352 0.00203 0.04608 0.01086 3.32766 0.00010 0.01074 0.00457 0.00088 3.28029 0.62981 0.92366 7.67402 0.00014 0.01272

47070

78353

139867

1.87425 2.08731

2.51416

59 1290 18659 33 716 234 66 160 365 7291 18 3905 1182 5963 13 31593 71 1380 50082 12736 7 — 21 1065

169 1744 42089 45 426 532 43 320 403 10325 18 6731 1885 10196 6 69220 105 2509 80441 33201 13 — 22 2819

333 2214 81146 66 372 1046 37 502 494 13328 19 8751 2186 28094 14 121775 145 4558 121451 58583 11 — 30 9540

0.00237 0.05135 0.74298 0.00131 0.02853 0.00933 0.00264 0.00638 0.01454 0.29031 0.00071 0.15551 0.04706 0.23743 0.00053 1.25798 0.00284 0.05495 1.99416 0.50711 0.00026 — 0.00084 0.04241

0.00598 0.03980 1.45863 0.00118 0.00668 0.01879 0.00067 0.00902 0.00887 0.23957 0.00034 0.15729 0.03929 0.50500 0.00025 2.18894 0.00261 0.08192 2.18313 1.05305 0.00019 — 0.00053 0.17148

0.00451 0.04646 1.12123 0.00120 0.01136 0.01418 0.00115 0.00853 0.01073 0.27504 0.00048 0.17932 0.05021 0.27162 0.00017 1.84399 0.00280 0.06683 2.14293 0.88447 0.00035 — 0.00058 0.07510

Marginalization and World Trade Appendix 7.19. Average Share of Individual Countries in Exports of Commercial Services Exports of Commercial Services ($ Mill.)

Share (per cent)

avg. 85–89 avg. 1990–94 avg. 1995–2000 1985–89 1990–94 1995–2000 World Australia Austria Belgium-Luxembourg Canada Denmark Finland France Germany Greece Iceland Ireland Israel Italy Japan Netherlands New Zealand Norway Portugal South Africa Spain Sweden Switzerland United Kingdom United States Afghanistan Albania Algeria Angola Antigua and Barbuda Argentina Armenia Aruba Azerbaijan Bahamas Bahrain Bangladesh Barbados Belarus Belize Benin Bolivia Bosnia and Herzegovina Botswana Brazil Brunei Darussalam Bulgaria Burkina Faso Burundi Cambodia Cameroon Cape Verde Central African Republic

523406 6223 14298 15521 12949 7668 3185 47476 32646 3973 362 1879 3676 26225 29894 19091 2016 8746 2893 2183 20069 8500 13487 39318 86477 4 25 496 123 203 1768 7 221 78 1251 834 221 519 109 39 65 113 0 74 2180 216 1114 21 3 27 433 22 16

903144 11342 25717 28978 20470 13412 4578 71875 55879 7918 475 3669 5492 51414 48171 35099 2781 12340 5831 3245 30709 13879 20058 59235 157568 5 42 529 103 343 2688 11 523 130 1413 557 377 652 189 101 117 154 307 177 3941 409 947 38 7 48 365 32 17

1327809 17106 30744 36730 31239 16338 6676 82111 79600 12288 764 10818 9676 62115 64816 49504 4224 13986 8277 4884 47110 17815 25490 98503 234893 7 172 892 151 396 4181 98 840 238 1659 715 293 954 823 130 137 210 597 251 6458 410 1626 43 3 130 347 78 12

100 100 100 1.1890 1.2558 1.2883 2.7318 2.8475 2.3154 2.9653 3.2086 2.7662 2.4739 2.2665 2.3526 1.4651 1.4850 1.2305 0.6085 0.5069 0.5028 9.0705 7.9583 6.1839 6.2372 6.1872 5.9948 0.7590 0.8768 0.9254 0.0692 0.0526 0.0575 0.3589 0.4063 0.8147 0.7023 0.6081 0.7287 5.0105 5.6928 4.6780 5.7114 5.3337 4.8814 3.6475 3.8863 3.7282 0.3852 0.3080 0.3181 1.6709 1.3663 1.0533 0.5527 0.6456 0.6234 0.4172 0.3593 0.3678 3.8342 3.4002 3.5480 1.6241 1.5367 1.3417 2.5768 2.2210 1.9197 7.5120 6.5588 7.4184 16.5219 17.4466 17.6903 0.0008 0.0005 0.0005 0.0047 0.0047 0.0130 0.0948 0.0586 0.0671 0.0235 0.0114 0.0114 0.0387 0.0379 0.0298 0.3377 0.2976 0.3148 0.0013 0.0012 0.0074 0.0421 0.0579 0.0633 0.0150 0.0144 0.0179 0.2390 0.1564 0.1250 0.1593 0.0617 0.0538 0.0423 0.0418 0.0221 0.0992 0.0722 0.0719 0.0209 0.0210 0.0620 0.0075 0.0112 0.0098 0.0124 0.0130 0.0103 0.0216 0.0170 0.0158 0.0000 0.0340 0.0449 0.0142 0.0196 0.0189 0.4164 0.4363 0.4864 0.0412 0.0453 0.0309 0.2129 0.1049 0.1225 0.0040 0.0043 0.0033 0.0006 0.0008 0.0002 0.0051 0.0053 0.0098 0.0827 0.0404 0.0261 0.0042 0.0036 0.0059 0.0031 0.0018 0.0009 (Continued )

261

The Implications of Declining Commodity Prices Appendix 7.19. (Continued ) Exports of Commercial Services ($ Mill.)

Share (per cent)

avg. 85–89 avg. 1990–94 avg. 1995–2000 1985–89 1990–94 1995–2000 Chad Chile China Colombia Congo Congo, Dem. Rep. Costa Rica ˆ te d’Ivoire Co Croatia Cuba Cyprus Djibouti Dominica Dominican Republic Ecuador Egypt El Salvador Equatorial Guinea Estonia Ethiopia Fiji Gabon Gambia Georgia Ghana Guinea Guinea-Bissau Hong Kong, China Hungary India Indonesia Iran, Islamic Rep. of Jordan Kazakhstan Kenya Kiribati Korea, Rep. of Kuwait Kyrgyzstan Lao People’s Dem. Rep. Latvia Lesotho Liberia Libyan Arab Jamahiriya Lithuania Macau, China Madagascar Malawi Malaysia Maldives Mali Malta Mauritania Mauritius

262

19 1036 3989 1132 68 191 348 394 0 448 999 21 17 830 415 3458 246 5 124 222 228 151 39 119 60 48 2 11761 921 3497 1199 254 1224 263 528 5 6230 934 4 11 184 21 31 79 124 786 81 29 2193 69 57 484 20 258

19 2267 9813 1800 55 145 844 445 1388 802 1998 34 42 1368 572 6555 310 6 280 247 420 261 65 204 110 64 6 24365 2812 5094 3452 548 1460 424 767 13 11607 1059 13 35 401 32 41 56 216 2138 147 31 5686 143 61 846 15 553

30 3736 23949 1925 90 133 1267 468 3588 2324 2726 27 79 2446 679 8872 492 7 1290 346 540 211 92 250 301 48 5 37782 5736 10935 5405 1001 1726 813 735 19 24954 1417 48 92 1030 44 46 32 920 2950 264 35 13125 307 65 1101 25 938

0.0036 0.1979 0.7620 0.2164 0.0130 0.0365 0.0666 0.0754 0.0000 0.0856 0.1908 0.0040 0.0032 0.1586 0.0794 0.6606 0.0469 0.0010 0.0236 0.0424 0.0436 0.0289 0.0075 0.0227 0.0114 0.0091 0.0003 2.2470 0.1759 0.6682 0.2292 0.0485 0.2338 0.0502 0.1008 0.0010 1.1902 0.1785 0.0008 0.0020 0.0352 0.0040 0.0059 0.0151 0.0237 0.1501 0.0154 0.0055 0.4189 0.0132 0.0109 0.0925 0.0039 0.0493

0.0021 0.2510 1.0866 0.1993 0.0060 0.0161 0.0935 0.0493 0.1537 0.0888 0.2212 0.0037 0.0046 0.1514 0.0633 0.7258 0.0344 0.0007 0.0310 0.0273 0.0465 0.0289 0.0072 0.0225 0.0122 0.0071 0.0007 2.6978 0.3113 0.5641 0.3822 0.0607 0.1617 0.0470 0.0850 0.0014 1.2852 0.1173 0.0014 0.0039 0.0444 0.0035 0.0045 0.0062 0.0240 0.2367 0.0162 0.0035 0.6295 0.0158 0.0067 0.0937 0.0017 0.0612

0.0022 0.2814 1.8036 0.1450 0.0068 0.0100 0.0954 0.0353 0.2702 0.1750 0.2053 0.0020 0.0059 0.1842 0.0511 0.6682 0.0371 0.0005 0.0971 0.0260 0.0407 0.0159 0.0069 0.0188 0.0227 0.0036 0.0004 2.8454 0.4320 0.8235 0.4070 0.0754 0.1300 0.0612 0.0553 0.0014 1.8793 0.1067 0.0036 0.0069 0.0776 0.0033 0.0035 0.0024 0.0693 0.2221 0.0199 0.0027 0.9885 0.0231 0.0049 0.0829 0.0019 0.0706

Marginalization and World Trade Moldova, Rep. of Mongolia Morocco Mozambique Myanmar Namibia Nepal Niger Nigeria Oman Pakistan Papua New Guinea Paraguay Peru Philippines Poland Qatar Romania Russian Fed. Rwanda Samoa Sao Tome and Principe Saudi Arabia Senegal Seychelles Sierra Leone Singapore Slovak Rep. Slovenia Solomon Islands Somalia Sri Lanka St Kitts and Nevis St Lucia St Vincent and the Grenadines Sudan Suriname Swaziland Syrian Arab Republic Taipei, Chinese Tajikistan Tanzania, United Rep. of TFYR Macedonia Thailand Togo Tonga Trinidad and Tobago Tunisia Turkey Turkmenistan Uganda Ukraine United Arab Emirates Uruguay Uzbekistan Vanuatu Venezuela Vietnam Yemen Yugoslavia Zambia Zimbabwe

17 75 1284 78 60 68 138 24 341 22 795 87 232 692 2111 2402 248 776 4481 32 19 2 2697 250 128 17 6411 0 0 15 2 295 38 100 31 183 64 49 528 4600 25 109 0 3374 84 11 246 1202 4293 39 7 1524 1367 388 143 34 821 119 0 4363 53 194

29 37 1813 157 148 167 278 17 888 34 1310 267 402 780 4428 4503 354 754 7460 24 33 4 3207 313 177 53 16797 832 883 30 0 580 75 190 53 78 43 96 1138 10362 41 241 72 8978 80 13 365 1779 9310 61 43 2450 2481 843 246 63 1232 682 106 1768 75 297

141 62 2478 280 481 341 561 12 822 225 1379 333 584 1404 8960 9794 574 1467 11501 28 57 9 4234 346 254 76 26622 2161 1998 51 0 856 92 293 103 45 85 107 1472 17020 57 569 177 14729 69 15 513 2571 17524 189 160 4029 3459 1330 332 95 1306 2455 149 0 77 635

0.0032 0.0143 0.2454 0.0148 0.0114 0.0130 0.0263 0.0046 0.0652 0.0043 0.1520 0.0167 0.0443 0.1321 0.4033 0.4588 0.0475 0.1483 0.8560 0.0062 0.0036 0.0003 0.5153 0.0477 0.0245 0.0033 1.2249 0.0000 0.0000 0.0029 0.0003 0.0563 0.0073 0.0191 0.0059 0.0349 0.0122 0.0094 0.1010 0.8789 0.0047 0.0209 0.0000 0.6445 0.0160 0.0022 0.0469 0.2297 0.8201 0.0074 0.0013 0.2911 0.2611 0.0741 0.0273 0.0065 0.1569 0.0228 0.0000 0.8337 0.0101 0.0371

0.0032 0.0041 0.2007 0.0174 0.0164 0.0185 0.0307 0.0019 0.0983 0.0037 0.1451 0.0296 0.0445 0.0864 0.4903 0.4986 0.0392 0.0835 0.8260 0.0027 0.0037 0.0005 0.3551 0.0347 0.0196 0.0058 1.8599 0.0921 0.0978 0.0033 0.0000 0.0642 0.0083 0.0211 0.0058 0.0086 0.0048 0.0106 0.1260 1.1473 0.0046 0.0267 0.0080 0.9940 0.0089 0.0014 0.0404 0.1970 1.0308 0.0068 0.0047 0.2713 0.2747 0.0934 0.0273 0.0070 0.1364 0.0755 0.0117 0.1957 0.0083 0.0329

0.0106 0.0046 0.1866 0.0211 0.0362 0.0257 0.0422 0.0009 0.0619 0.0169 0.1039 0.0251 0.0440 0.1058 0.6748 0.7376 0.0432 0.1105 0.8661 0.0021 0.0043 0.0007 0.3189 0.0260 0.0191 0.0057 2.0050 0.1628 0.1505 0.0038 0.0000 0.0645 0.0069 0.0221 0.0078 0.0034 0.0064 0.0080 0.1109 1.2818 0.0043 0.0428 0.0133 1.1093 0.0052 0.0011 0.0386 0.1936 1.3198 0.0142 0.0120 0.3034 0.2605 0.1002 0.0250 0.0072 0.0983 0.1849 0.0112 0.0000 0.0058 0.0478

263

The Implications of Declining Commodity Prices Lao PDR

LDCs: Net Shifts in 1995–2000 in comparison with 1980–85

700

Myanmar

300 200

Tanzania

Guinea

Senegal Togo Congo, Dem. Rep.of

Central African Rep.

Ethiopia Mauritania Malawi

Madagascar

Sudan

Somalia Zambia Liberia

Burundi Haiti Niger

−400

Rwanda

−300

Afghanistan Sierra Leone

−200

Mozambique Uganda Angola

0

−100

Chad Guinea-Bissau Mali

100

Bangladesh Bhutan

400

Nepal

500

Burkina Faso Yemen, Republic of Benin

600

Appendix 7.20. Average net shift in exports of individual LDCs in 1995–2000 in comparison with 1980–85

Maldives

Mauritius

Dominica

Malta

Botswana

Lesotho

Cyprus

Seychelles

St Lucia

Grenada

Cape Verde

St Kitts and Nevis

Swaziland

Samoa

Comoros

Papua New Guinea

Jamaica

Belize

Solomon Islands

Guyana

Fiji

Kiribati

Barbados

Vanuatu

Tonga

Djibouti

Gambia

Sao Tome and Principe

0.00

Gabon

50.00

Bahrain

100.00

Suriname

per cent of average 1980–85 exports

150.00

Trinidad and Tobago

200.00

St Vincent and the Grenadines

250.00

Antigua and Barbuda

Note: Cambodia was found have the highest average net positive shifts in 1995–2000. On the basis of its 1980–85 average share in world exports, Cambodia’s average exports for 1995–2000 were predicted to be about $65 million. In reality, its exports in the late 1990s averaged at $806 million resulting in a net positive gain of 3807 per cent. Since gains/losses of individual countries are measured on a single scale in the ﬁgure, inclusion of Cambodia make inter-country graphical comparison for others indistinct.

−50.00 −100.00 −150.00 −200.00 −250.00

Appendix 7.21. Average net shift in exports of small states in 1995–2000 in comparison with 1980–85

264

Marginalization and World Trade Relationship Between GDP and Export Growth Rates LDCs 12

GDP growth rate (per cent)

10

y = 0.3043x + 1.0236 R 2 = 0.4734 Maldives

8 Solomon Islands 6

Bhutan

Cape Verde

4

Gambia

2 Djibouti

0 DR Congo 0

2 4 6 8 Exports growth rate (per cent)

10

12

14

16

Appendix 7.22. Exports and GDP Growth Rates in LDCs

Relationship Between GDP and Export Growth Rates LDCs 12 y = 0.3943x + 1.1521 R 2 = 0.4537

GDP growth rate (per cent)

10

Botswana Maldives

8 Cape Verde 6

St Kitts and Nevis

4

Gambia

2 Djibouti 0 0

Suriname 2 4

6 8 10 12 Exports growth rate (per cent)

14

16

18

Appendix 7.23. Exports and GDP Growth Rates in LDCs

265

This page intentionally left blank

PART III Mitigating the Impacts for Commodity Dependent Countries

This page intentionally left blank

8 Instruments for Addressing Commodity Price Behaviour Mohammad A. Razzaque, Philip Osafa-Kwaako, and Roman Grynberg

While volatility in commodity prices triggers the problem of export earnings instability, the long-term trend decline in relative prices has caused sustained foreign exchange losses for commodity-dependent LDCs, SVs, and HIPCs. There have been a number of initiatives at the international level either to ensure price stability in commodity markets or to help producing countries maintain export earnings stability. The basic objective of this chapter is to provide a brief review of the instruments employed to intervene in commodity markets, as well as of other support mechanisms which deal with export shocks in poor countries, and to examine whether they are useful in dealing with the secular decline in the relative prices of primary commodities. The idea of intervening in the operations of commodity markets can be traced to the ﬁrst half of the twentieth century, when British and Dutch ofﬁcials employed export restraints under various international agreements to regulate prices of rubber, tea, tin, sugar and wheat produced in their colonies (Herrmann et al., 1993). Following World War II, private stock-holding in Britain proved inadequate in reducing commodity price volatility, and Lord Keynes proposed the creation of international buffer stock arrangements to reduce price ﬂuctuations in commodity markets. What was envisaged was the creation of a third Bretton Woods institution, ‘Commod Control’, to stabilize world commodity prices. Keynes’ scheme was never put into practice, but other intervention tools have been employed as part of international commodity policy. For the post-war period, one can identify four main instruments used in international commodity policy: 1. International commodity agreements (ICAs); 2. External Compensatory Finance;

269

Mitigating the Impacts of Dependent Countries 3. Preferential trade arrangements; 4. Market-based commodity risk management instruments. Of these, market-based commodity risk instruments are relatively recent, and are currently being considered and experimented with. The three other instruments have a long history of international policy negotiations, which have been closely scrutinized in the economic literature.1 In general, all the instruments were designed to provide export earnings stabilization, but some of the ICAs also implicitly raised absolute price levels. Addressing the problem of secular decline in relative prices was never an explicit objective of any of the schemes; they have generally focused on price volatility. The majority of the instruments are now effectively defunct or under-utilized, so this review is essentially historical.

8.1. International Commodity Agreements Supply management of commodities in international markets was a dominant strategy in the management of international commodity prices in the second half of the twentieth century. The objective was to stabilize prices by the use of export quotas or buffer stock (national or international) arrangements. Most commodity-producing countries maintained individual reserves but these hardly affected world market prices. Having witnessed the success of OPEC countries in controlling world oil prices, developing countries realized the beneﬁts of stronger international producer associations in securing favourable commodity prices. Following the speciﬁcations of the Havana Charter in 1948, ICAs were to be negotiated between producing countries and consumers, with agreements lasting no more than ﬁve years, after which repeated renewals were expected.2 Agreements were initially designed to mitigate unexpected price ﬂuctuations and ensure long-run equilibrium between forces of supply and demand, thus precluding the raising of prices above market trends. With support from UNCTAD, various international commodity agreements were negotiated, renewed or strengthened—the most prominent ones being the International Sugar Agreement (ISA), the International Tin Agreement 1 More detailed historical overviews are provided by World Bank (1999) and Maizels (1992). Other programmes proposed in the past to provide supplementary ﬁnance include the Olano Mutual Insurance Scheme (1953), Development Insurance Fund (1961), the Organization of American States Proposal (1962) and proposals by various national governments—France (1963), USA (1975), Sweden (1977), and Germany (1978). 2 The Havana Charter established the Interim Coordinating Committee for International Commodity Arrangements (ICCICA), under which agreements for coffee, sugar, tin and wheat were established. ICCICA was subsequently absorbed into UNCTAD in 1964.

270

Table 8.1. Salient Features of Five Important International Commodity Agreements

ICAs

Time of Establishment

Mode of Operation

Main Problems

Time of Suspension/ Collapse

International Sugar Agreement (ISA)

1954

Export controls

The ﬁrst major problem occurred in 1962 when supplying countries were unwilling to accept Cuba’s demand for increased quota allocation as it faced an export restriction in the US market. This resulted in the collapse of the Agreement in 1963. ISA was later renegotiated. In the 1980s the EU emerged as the single largest net exporter of sugar and the US also supported domestic production with a stringent import quota regime. Since the EU and the US remained outside the cartel, the ISO had very little capacity to inﬂuence prices.

1983

International Tin Agreement (ITA)

1954

Buffer stock and export controls

In the early 1980s the US failed to renew its membership of the ITA, triggering a situation where there was insufﬁcient ﬁnance to hold stock. Futures market operation in the London Metal Trade became very expensive due to the depreciation of the US dollar, making the International Tin Council insolvent. At the same time increased supply from low-cost non-members, such as Brazil, combined with increased competition from aluminium to cause a rapid fall in prices.

1985

International Coffee Agreement (ICoA)

1962

Export controls

Despite ICoA’s success in maintaining high and stable prices, there were disagreements among members over distribution of quotas. Consumers’ taste shifted to ‘arabica beans’ and other ‘milds’ but the quota allocation favoured large robusta producers, such as Brazil. Because of political changes, Brazil, which played a central role in the operation of ICoA, lacked a clear coffee policy and therefore failed to enforce an agreement on other producing countries, causing the break-up of ICoA.

1989

International Cocoa Agreement (ICCA)

1972

Buffer stock

During the ﬁrst two phases of the Agreement prices were maintained above the ˆ te ceiling, encouraging excessive planting in non-member countries such as Co d’Ivoire, Brazil, Indonesia and Malaysia, which resulted in prices lower than the stabilization range in the next two rounds of ICCA. The third and fourth rounds of ICCA were chronically under-ﬁnanced. Total stock at the end of the 1986–87 crop year was 650,000 tons, but a buffer stock of 150,000 tons only could be operated, making the agreement ineffective in inﬂuencing prices. The low marginal cost of harvesting beans and the long life of trees make the management of cocoa production and buffer stock difﬁcult.

1988

International Natural Rubber Agreement (INRA)

1980

Buffer stock

The Agreement attempted an active market intervention in terms of a daily market indicator price, which was the average of the Kuala Lumpur, London, New York and Singapore cash prices. Producing countries disagreed on the use of rounded versus unrounded price. The price range of intervention also created tensions among members. Finally, the two main producers, Malaysia and Thailand, withdrew in 1999, leading to the transformation of the agreement into a mere study group.

1999

Source: Compiled from Gilbert (1987 and 1996) and Page and Hewitt (2001).

Mitigating the Impacts of Dependent Countries (ITA), the International Coffee Agreement (ICoA), the International Cocoa Agreement (ICCA) and the International Natural Rubber Agreement (INRA). ICAs, as employed in international commodity policy, aimed at achieving the two simultaneous goals of reducing variability in prices and raising depressed price levels. An Integrated Programme for Commodities (IPC) was negotiated during UNCTAD IV in 1976. The IPC envisaged the establishment of a Common Fund for Commodities (CFC) as an umbrella institution for negotiating additional commodity agreements, and providing ﬁnancial economies of scale compared with the establishment of separate commodity funds. The CFC had two accounts—the ﬁrst aimed at ﬁnancing buffer stocks, and the second promoting research on crop production techniques. The two accounts of the CFC became active only in the late 1980s. By the late 1990s, most commodity agreements had disintegrated, as a result of unfavourable market conditions and, in some cases, disagreement among member countries (see Table 8.1).3 The breakdown of the sugar and cocoa agreements was triggered by unfavourable market conditions, while disagreement among producing members ended the coffee and rubber agreements. At present, most international commodity organizations no longer attempt to intervene in the markets with a view to inﬂuencing prices or supplies, i.e. even price stabilization measures have been abandoned. A large number of them have been converted into ‘study groups’ providing market information on production and prices. The CFC has been transformed into a grant-making institution, primarily supporting agronomic research on individual commodities (CFC, 2002).

8.2. External Compensatory Finance In the post-war period, shortfalls in export earnings threatened to disrupt the development plans envisaged by many developing countries. Supplementary ﬁnance was conceived in response to this, which aimed at supporting export earnings stabilization of commodity-dependent poor countries. Direct external compensation was an attractive policy tool, in contrast to other commodity market instruments, as it provided the least distortionary means of intervention. The two main schemes of this type are: 1. IMF Compensatory Financing Facility (CFF) 2. EC-ACP Programmes: STABEX, SYSMIN, and COMPEX 3 A full discussion of the operation and breakdown of commodity agreements is provided by Gilbert (1996). The spectacular collapse of the International Tin Agreement is also discussed in Andersen and Gilbert (1988).

272

Instruments for Addressing Commodity Price Behaviour

8.2.1. Compensatory (and contingency) ﬁnancing facility (CCFF) Established in 1963, the IMF Compensatory (and Contingency) Financing Facility is the sole initiative undertaken by Bretton Woods institutions to provide some form of compensatory ﬁnance. The facility operates as a concessional loan arrangement that helps recipient countries to mitigate balance of payments difﬁculties in the face of export earnings shortfalls caused by unexpected exogenous factors such as unfavourable movements in prices and adverse production shocks (for example drought). The CCFF considers compensation for net export earnings, rather than on a commodity-by-commodity basis and consequently captures portfolio effects present in a country’s entire export basket. Since its inception, the ﬁnancial architecture of the CCFF has been subject to periodic modiﬁcation. The credit quota for members was increased from 25 per cent in 1963 to 75 per cent in 1975. The deﬁnition of balance of payment shortfalls has also been frequently revised. In 1979, compensation was extended beyond traditional export earnings to include deﬁcits arising from low receipts from tourism or workers’ remittances. Coverage was further extended to include compensation for increases in cereal import costs in 1981 and, ﬁnally, to include all services where adequate data were available (World Bank, 1999). A signiﬁcant modiﬁcation was made to the scheme in 1987 to include a contingency facility that permitted member countries to borrow, ex ante, in anticipation of balance of payment shortfalls. The new Compensatory and Contingency Financing Facility provided up to 122 per cent of national quota drawings, but required recipient countries to agree to a series of IMF-led programmes aimed at addressing balance of payments problems.4 In recent years, however, ﬁnancing from the CCFF has rarely been used by LDCs and HIPCs, partly because of the high rates of interest charged (about 3.52 per cent) compared with other IMF sources of credit, such as the Poverty Reduction Growth Facility (PRGF). For most of the 1990s, credit facilities from the CCFF have been used by a few countries, such as Russia and Algeria, to support their crude oil export earnings.

8.2.2. EC-ACP programmes: STABEX, SYSMIN and COMPEX STABEX (the French acronym for STABilization des recettes d’EXportation) was established in 1975 by the European Community (EC) as part of the Lome´ arrangements with ACP member countries, and has served as a major delivery mechanism for EC aid (about one-eighth of the EU aid budget). STABEX was 4 Speciﬁc ﬁnancing available is distributed as: 20 per cent of the national quota for export shortfall, 20 per cent of the quota for external contingency and 10 per cent of the quota for import costs of cereal, plus an optional 15 per cent.

273

Mitigating the Impacts of Dependent Countries initially conceived as a concessional loan agreement, but evolved into a grantbased program.5 The STABEX scheme was essentially in operation under the four Lome´ Conventions spanning 1975–2000 and was discontinued under the Cotonou Agreement. As an aid instrument, STABEX provided periodic export earnings compensation to many ACP countries, disbursing a total ECU4.4 billion over the ˆ te d’Ivoire and Senegal being major beneﬁciaries period 1975–1998, with Co (World Bank, 1999). The required trigger was a loss of export earnings (to the EU) relative to a six-year trend. STABEX funds were relatively small: the most recent funding ceiling of ECU1.8 billion over the ﬁve-year period 1996–2000 implied an average disbursement of ECU360 million per annum. This sum represented the net resources for which some ﬁfty or so states were eligible. It is important to note that STABEX focused on a restricted set of countries, and provided compensation for only a limited number of agricultural exports supplied to EU markets. In Lome´ IV an ‘all destinations’ derogation was granted to certain ACP countries exporting commodities to countries other than the EU. Under the scheme, a total of 50 raw and processed agricultural, ﬁshery and forestry products were covered for export compensation, with nearly 80 per cent of transfers made to a shortlist of commodities: coffee, cocoa, groundnuts, cotton, and copra. The STABEX scheme therefore provided compensation on a commodity-by-commodity basis, in contrast to the CCFF which provided general compensation for shortfalls in net export earnings. In the course of its operation, the STABEX scheme was criticized for two main reasons. First, signiﬁcant time lags often existed between approval of compensation for member countries and the actual disbursement of funds. These frequent delays in delivering supplementary ﬁnance implied that STABEX disbursements tended to be pro-cyclical, rather than responding countercyclically to adverse shocks (Herrman et al., 1993). Second, there have been marked shifts in policy conditionalities attached to STABEX transfers (Hewitt, 1993). Initial compensation under the scheme was untied, and provided extra budgetary support to recipient governments. However, by the 1980s, the STABEX scheme was projectized, with strict conditionalities imposed on the use of transfers, discouraged diversiﬁcation and required programmes to promote recovery of the particular commodity in distress (at least under articles of Lome´ IV). This tying of aid to the recovery of ailing commodities was prone to the fallacy of composition argument, as it encouraged further dependence on weak commodities, the prices of which were already in secular decline. 5 Original articles under Lome´ required designated funds as grants for least developed countries in the ACP, repayable under concessional terms by other member states. However, in 1991, outstanding reimbursements due from Senegal, Gabon, Madagascar, and Jamaica were cancelled; in return, ACP countries were required not to pursue compensation for outstanding STABEX claims from 1980–1 and 1987–8.

274

Instruments for Addressing Commodity Price Behaviour By the mid-1980s, following demands by commodity-dependent countries not adequately covered by STABEX transfers, two other complementary programs were created by EC donors—SYSMIN and COMPEX. SYSMIN was established during Lome´ II (1980–85) and was designed to focus on minerals and mining (in contrast to the predominant focus on agriculture in STABEX). The speciﬁc list of commodities covered included copper, cobalt, phosphates, manganese, bauxite, alumina, tin, iron ore, uranium, and gold. SYSMIN receipts were intended to be partly supplementary ﬁnancing, and also to support technical and ﬁnancial assistance for exploratory (geological and mining) research. In the period 1995–2000, ECU575 million was allocated, and the scheme was expected to have transferred a total of ECU1.7 billion over its 20-year lifespan. Transfers of SYSMIN funds were triggered either by adverse external shocks which threatened the viability of important components of the mining industry or by shortfalls in export earnings which were likely to derail existing development projects. The effectiveness of SYSMIN aid has been questioned, as transfers were often made to corrupt governments or to mining sectors run by foreign organizations which often did not need external aid.6 COMPEX emerged in the late 1980s, and was established to extend the beneﬁts of STABEX to non-ACP LDCs. The scheme proved to be largely symbolic with very few states beneﬁting. With the EC lacking the necessary administrative structure, and the absence of a treaty document, it was difﬁcult for LDCs to request funding (Page and Hewitt, 2001). In the current review, it is found that compensatory payments are mostly non-existent. Existing EU-ACP agreements within the Cotonou framework are vague about the establishment of other compensatory ﬁnance schemes. Article 68 of the Cotonou Agreement recognizes the problem of commodity dependence, and aims to ‘mitigate the adverse effects of any instability in export earnings’. Consequently, a new scheme, FLEX (an acronym for FLuctuations in EXport earnings), was suggested under the EU-ACP Cotonou Agreement.

8.3. Preferential Trade Agreements A third channel through which donors have indirectly supported commodity prices has been through the use of various preferential trade arrangements. The case of sugar under EU-ACP trade cooperation is the best example. Under the terms of the Lome´ Agreement, a scheme granting preferential prices for sugar supplied by the ACP countries was established. A number of ACP countries received special quota arrangements that ensured preferential prices 6 Page and Hewitt (2001) argue that there is inconclusive evidence that SYSMIN transfers supported weak mining sectors in Zambia and the Democratic Republic of Congo.

275

Mitigating the Impacts of Dependent Countries for their sugar exports, equivalent to domestic sugar beet prices in the EU. Apart from sugar, commodity protocols also operated for bananas, beef/veal, and rum.7 For those ACP countries that received quotas, the commodity protocols have proved to be an invaluable source of foreign exchange earnings from the commodities in question. Moreover, given that the commodity arrangements linked export prices to domestic EU intervention prices, this provided high and stable prices for ACP farmers. However, given the expansion of the EU, the possible results of the Doha Round and the internal pressures for reform of the EU Common Agricultural Policy, these arrangements will be phased out by the end of the current decade, leaving affected ACP states to make substantial market adjustments. For developing countries that have not been beneﬁciaries of the commodity protocols, the overall economic impact of EU preferential pricing arrangements is mixed.

8.4. Market-Based Instruments for Commodity Risk Management and Insurance Schemes At present, signiﬁcant donor attention (including from the World Bank and the EU) is aimed at providing ﬁnancial market derivatives which assist developing countries to manage commodity price risks. Smoothing of income streams may be achieved using a number of tools, including futures, options, swaps and commodity-linked notes.8 Unlike ICAs or other compensatory ﬁnancing schemes, commodity risk management tools do not intend to provide external resources to stabilize export earnings. In addition, under these schemes price risks are envisaged as increasingly allocated to private traders and farmers rather than absorbed by governments. However, since farmers do not generally have direct access to these instruments, the role of intermediaries is crucial in the functioning of the schemes.9 The goal of managing risk and liability in commodity-dependent countries is, however, fraught with difﬁculty. Such difﬁculties include the needed regulation reform, the identiﬁcation of appropriate intermediaries, developing local trading exchanges, and the ability of the emerging private sector to make full use of the available range of modern commodity marketing, price 7 Only the protocols for sugar and beef/veal provided a guaranteed price for ﬁxed quantities of exports from beneﬁciary countries. The banana, rum and sugar protocols granted duty-free access to ﬁxed quantities supplied by ACP countries, while the arrangement for beef/veal considered guaranteed price, subject to a reduced level of duty, for a pre-determined quantity. 8 UNCTAD (1999) and Page and Hewitt (2001) provide discussions on these instruments. 9 Large private traders and banks are usually considered to be in the best position to act as intermediaries.

276

Instruments for Addressing Commodity Price Behaviour risk management and ﬁnancing instruments. Donor funding may be relevant in this regard in paying insurance premiums and covering producers in developing countries for their exposure to commodity-related risks. Tools like insurance schemes have also been proposed for ex ante price risk management. In 1999, the World Bank (ITF) considered a market-based international commodity price insurance mechanism consisting of price ﬂoor guarantees for producers/exporters and price ceiling guarantees for consumers/ importers. Under the scheme it is thought that international intermediaries would bridge the gap between private providers of insurance and other parties in developing countries. It would also provide local entities with the core services and technical assistance needed to extend the market to them. Market-based instruments for commodity risk management and insurance schemes are, however, concerned solely with the problem of revenue ﬂuctuations due to commodity price volatility, clearly distinguishing this goal from that of addressing long-term price weaknesses, as ITF (1999: 17) observes: The scheme focuses on short-term price ﬂuctuations, not price trends . . . The negative impact of the long-term deterioration of commodity terms of trade faced by many developing countries, needs to be dealt with by broader macroeconomic policies and development strategies. (ITF, 1999:17)

8.5. Conclusion From the above discussion, it can be concluded that there does not currently exist any policy instrument that explicitly addresses the problem of weakness in relative commodity prices. While price stabilization was the principal motive of the international commodity agreements, they also made some attempt to raise depressed price levels absolutely. However, no consideration was made of the trends in real prices (i.e. commodity prices relative to manufactured goods). On the other hand, external compensatory ﬁnancing mechanisms (such as IMF, CFF, and STABEX) focused only on the shortfalls in absolute export earnings; export earnings from commodities and commodity prices were not speciﬁcally targeted. Various commodity protocols under EU-ACP trade arrangements guaranteed preferential prices for speciﬁc commodities. This has eased the problem of low prices of relevant commodities supplied by the recipient countries. The coverage of commodities in these arrangements has been quite limited. Finally, the relatively recent attempts with market-based instruments and insurance programmes are not designed to address the problem of trends in real prices. Rather, these schemes are intended to smooth out price ﬂuctuations through the participation of the private sector.

277

9 Commodity Prices and the Debt Relief Initiative Mohammad A. Razzaque, Philip Osafa-Kwaako, and Roman Grynberg

Over the past two decades a number of low-income countries have amassed large stocks of external debt. Among the primary reasons usually cited for the debt crisis are the economic shocks of the 1970s, declining terms of trade, highly volatile and declining commodity prices, heavy reliance on foreign aid and borrowing, and weak ﬁscal management and governance.1 The lack of diversiﬁcation in export structure, together with excessive dependence on commodities facing price and income inelastic demand in the world market, has accentuated the process of accumulation of debt in an overwhelming majority of the heavily indebted poor countries. In fact, these countries are eventually caught in a vicious circle: commodity price declines exerting terms of trade shocks causing reduced export earnings and hence reduced capacity to import, leading to lower economic growth and necessitating more and more borrowing.2 The late 1990s witnessed some genuine and serious efforts to mitigate the debt problems of these countries. This section of the study argues that while the latest round of debt relief initiatives, known as the Enhanced HIPC Initiative, is laudable, the crucial factors of secular decline and sharp movements in commodity prices pushing most poor countries into the debt trap have not been adequately taken into

1 It should be noted here that the debt problem of heavily indebted poor countries is quite different from that of the middle-income countries of Latin America. In most cases, lowincome countries borrowed from bilateral donors and multilateral agencies such as the World Bank and IMF at concessional rates of interest. On the other hand, the middle-income countries borrowed mainly from the private commercial ﬁnancial institutions at market interest rates. 2 Despite implementing extensive policy reforms to overcome the problem of internal and external imbalances, rates of external indebtedness continued to increase in the countries labelled as the highly indebted poor countries.

278

Commodity Prices and Debt Relief consideration. This omission threatens the credibility of the initiative if the poverty reduction strategies formulated to enable countries to qualify for the assistance fail to energize exports in poor countries. We emphasize the importance of establishing a support mechanism for the HIPC countries so that, at least in the short run, they are covered against falling commodity prices while implementing their poverty reduction strategies. In the long run, however, their permanent exit from the debt problem will depend on their success in diversifying their export structures and achieving a more favourable share of international trade and ﬁnance.

9.1. Commodity Prices and Debt The overall debt of low-income poor countries increased substantially in the 1980s and 1990s. In 1970 the total debt stock of the 42 heavily indebted poor countries stood at US$6.7 billion, rising to US$59 billion over the following 10 years.3 It was in the 1980s that debt stocks exploded with a volume of outstanding payments of US$190 billion. Since by the late 1980s the situation for HIPCs had already become unsustainable with persistent debt-overhang, in the last decade there has been only a small increase of about US$15 billion in the overall debt burden.4 As a proportion of GDP, the total outstanding debt of HIPCs in the 1990s reached 95 per cent, in comparison with about 35 per cent for other lower-middle-income countries. The rapid and concomitant accumulation of arrears of low-income countries undoubtedly reﬂects their inability to make regular debt service payments. Weak macroeconomic management of the economy supposedly affecting growth performance, together with the inability to make necessary ﬁscal adjustments in the face of exogenous shocks, is often given much emphasis in explaining the debt default phenomenon.5 However, it cannot be 3 Total external debt here includes both long-term and short-term outstanding loans together with the use of IMF credit. 4 Also, in the 1990s under various initiatives the low-income countries received some debt forgiveness extinguishing a portion of their outstanding debt stock. 5 According to Easterly (2001), the growth slow-down played an important role in the debt crisis of HIPCs in the 1980s and 1990s. The author attributes the poor growth performance of HIPCs to lower investment in infrastructure, such as transport and communication, overvalued exchange rates, lower primary and secondary enrolments, and lower monetization in the economy as measured by the ratio of broad money to GDP. The terms of trade shock is, however, not considered and therefore its impact on economic growth is overlooked. There is evidence of declining commodity prices translating into terms of trade shocks (Grilli and Yang, 1988; Bleaney and Greenaway, 1993) resulting in an immediate negative and signiﬁcant impact on growth (Dehn, 2001; Dehn et al., 2003; Singer and Edstrom, 1993). For African countries, Deaton (1999) provides further stronger evidence of a positive relationship between commodity prices and growth: for a country whose commodity exports are a third of GDP, a commodity price increase of 1 per cent will directly increase national income by 1 per cent plus another half of 1 per cent. With regard to bad governance and increased government

279

Mitigating the Impacts of Dependent Countries overlooked that as most poor countries rely heavily on primary commodities for exports, the secular decline in commodity prices has certainly undermined their capacity to service debt.6 In all the World Bank country case studies on HIPCs, falling commodity prices have been identiﬁed as one of the principal reasons for the debt crisis (Gautam, 2003).7 Figure 9.1 presents the outstanding stock of debt of HIPCs (in real terms and measured on the left vertical axis) vis-a`-vis the relative price of commodities (measured on the right vertical axis). It can be seen that the mounting debts of HIPCs in the 1980s were accompanied by a sharp fall in the real commodity price index. A simple regression of the logarithm of real debt on a constant and the logarithm of relative commodity price showed that a 1 per cent fall in commodity price was associated with a 0.91 per cent increase in outstanding stock.8 The statistical association and Figure 9.1, therefore, seem strongly to suggest that commodity prices have been central in the debt crisis of HIPCs. In order better to appreciate the severity of the impact of the secular decline in commodity prices, we juxtapose the cumulative foreign exchange losses (1985–2000) for a number of countries from some primary commodities and the respective countries’ outstanding debt (Table 9.1).9 Note that the foreign exchange losses from the commodities as considered are not exhaustive for all the countries listed in the table; in most cases, countries will have other exportable primary commodities that have not been taken into account. expenditure followed by a positive shock in commodity prices, it must be acknowledged that the weak state of management is a general characteristic of poor countries. Nevertheless, it has also been argued that growth of public spending was not always reckless and in many cases concentrated on the development of infrastructure and other long-term investment (Geda, 2002). 6 Easterly (2002) ﬁnds the growth of terms of trade in HIPCs to be not signiﬁcantly different from that of a set non-HIPC developing countries, leading him to suggest that adverse movement in terms of trade cannot explain the debt problem. However, it should be understood that the classiﬁcation of a country as a non-HIPC does not necessarily mean that it does not have a debt problem. The criteria according to which countries are classiﬁed as HIPCs are not strictly scientiﬁc and the classiﬁcation is considered to be a ‘rule of thumb’ only. More importantly, similar movements in terms of trade of countries with a different composition of exports can have different consequences. For example, countries exporting only primary commodities will ﬁnd their revenues falling in the case of a decline in price as the demand for commodities is price inelastic in nature. Manufacturing exporters can have rising exports despite falling prices as they have a fairly elastic demand curve. 7 ˆ te d’Ivoire, EthiThe country case studies include Bolivia, Burkina Faso, Cameroon, Co opia, Ghana, Malawi, Mozambique, Tanzania, Togo, Uganda, Zambia, and Yemen. A summary of these case studies is given in Gautam (2003). 8 Due to the potential problem of the non-stationarity of data, the regression was estimated using the Phillips-Hansen Fully Modiﬁed Ordinary Least Squares (PHFMOLS) procedure. The PHFMOLS technique yields standard errors that are valid for inference and the resultant ‘t’-ratio on the relative commodity price conﬁrmed the statistical signiﬁcance of the elasticity coefﬁcient at less than the 1-per-cent level. Short size of the sample did not allow us to use the more powerful procedure of the Johansen cointegration test. 9 In Chapter 6 we estimated the cumulative foreign exchange losses in 1984–86 prices; but for comparing with outstanding debt stocks, they have been converted into 1999 dollar prices.

280

Commodity Prices and Debt Relief 170

140000 130000

150 130

US$ Million

110000 100000

110 90000 90

80000

Index: 1985=100

120000

70000 70 60000 Debt stock in 1985 Prices

Real Commodity Prices 1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

1984

1983

1982

1981

50 1980

50000

Figure 9.1. Real Commodity Price Index and Real Outstanding Debt in HIPCs Note : (1) Debt stocks are measured on the left vertical axis and the real commodity price index is on the right vertical axis. (2) The US consumers’ price index (CPI) has been used to convert debt stock in nominal US dollars into real terms. (3) The real commodity price index is the UNCTAD composite price index for primary goods relative to manufactures as used in Chapter 3.

Table 9.1. Foreign Exchange Losses from Commodities and Outstanding Debt

Countries Benin ˆ te d’Ivoire Co Ghana Honduras Mali Papua New Guinea Sao Tome and Principe Solomon Islands Togo Uganda Vietnam

Commodities considered Cotton Cocoa, Cotton, Rubber, Coffee, Palm Oil Cocoa Coffee Cotton Cocoa, Rubber, Palm Oil, Coconut Oil Cocoa Palm Oil Cotton Cotton, Coffee Rubber, Tea, Coffee, Coconut Oil

Outstanding Debt (US$ Million)

Cumulative Foreign Exchange Loss (1985–2000) (US$ Million)

Foreign Exchange Loss as % of debt stock

1589 11582

355 19456

22.3 167.9

6240 4366 2956 2500

6330 2482 248 1590

101.4 56.85 8.4 63.6

225

83

36.8

141

88

62.4

1485 3544 12578

262 2545 6004

17.6 71.8 47.7

Note: Cumulative foreign exchange losses in 1984–86 prices have been estimated in Chapter 4 but in the above table they have been adjusted upward to current prices in 1999. Source: The data on debt are in most cases from the UNCTAD Handbook of Statistics CD-ROM 2001 and represent the outstanding stock in 1999. Data from Global Development Finance (World Bank, 2003) have been used for 2001.

281

Mitigating the Impacts of Dependent Countries Nevertheless, the comparison of loss with the outstanding debt stock is striking. By the end of the 1990s, Ghana had accumulated a debt stock of US$6240 million, which appeared to be equivalent to the amount it had lost in foreign exchange as a result of the declining price of cocoa alone. The calculation of forgone foreign exchange covered a set of commodities that happened to ˆ te d’Ivoire (cocoa, coffee, comprise the ﬁve most important export items of Co cotton, rubber, and palm oil)—cumulative losses from which amount to 168 per cent of the country’s outstanding debt stock. In the case of Uganda, lost revenues from coffee and cotton equal about 72 per cent of its debt stock. Among others, the small state of Sao Tome and Principe, which acquired a debt stock-GNI ratio of 735 per cent (World Bank, 2003), lost an amount equivalent to 37 per cent of its arrears because of the falling cocoa price. On the whole, the information presented in Table 9.1 suggests that the loss in foreign exchange earnings because of secular decline in commodity prices has signiﬁcantly contributed to the debt problem of poor countries.

9.2. Debt Relief Initiatives There have been several attempts to relieve poor countries of their external debts (see Table 9.2). Traditional debt relief initiatives have included: (1) concessional ﬂows and rescheduling; (2) stock-of-debt operations; (3) bilateral forgiveness of ODA claims by Paris Club creditors; and (4) private commercial debt relief and buy-back operations (Daseking and Powell, 1999). These debt restructuring, rescheduling and forgiveness schemes have proved to be insufﬁcient to resolve the debt problem of poor countries. Indeed most donors have frequently resorted to a ‘defensive lending’ strategy—involving debt rescheduling and fresh loans and grants to prevent default by borrowing countries.10 In 1996 the World Bank and the IMF proposed the Heavily Indebted Poor Countries (HIPC) Initiative. Its main objective was to bring down the external debt burdens of the eligible countries to a ‘sustainable level’ and to eliminate the problem of debt overhang, thereby paving the way for a permanent exit from the process of debt rescheduling. In its design it was recognized that a precondition of any effective debt relief programme should be broad and equitable participation by all creditors—bilateral, multilateral, and commercial, and that the debt relief programme should be complemented by additional resource ﬂow to enable the poor countries to initiate an effective poverty reduction strategy. The Initiative was reviewed and enhanced (and

10 Despite the debt problem the net transfer in the form of fresh loans, grants and other transfers has been positive and relatively high in HIPC countries (Birdsall et al., 2002; Gautam, 2003). UNCTAD (2002a) shows that throughout the 1990s for the group of LDCs gross aid disbursements were strongly correlated with debt service payments.

282

Commodity Prices and Debt Relief Table 9.2. Debt Relief Initiatives Initiative

Main Feature

UNCTAD, 1977–79

Led to ofﬁcial creditors writing off US$6 billion in debt to 45 poor countries in terms of elimination of interest payments, the rescheduling of debt service, local cost assistance, untied compensatory aid and new grants to reimburse old aid.

Venice G7 Summit, 1987

Called for interest rate relief on debt of low-income countries.

Special Programme of Assistance (SPA) to Africa, 1987

Initiated by the World Bank to help African countries (African IDA-only borrowers with a ratio of debt service to exports above 30 per cent) to service their ofﬁcial debt. The IMF complemented the SPA with an enhanced structural adjustment facility. Agreed a menu of options including partial forgiveness, longer maturities and lower interest rates. The level of reduction was deﬁned as 33.33 per cent. World Bank and IMF initiatives to help mainly middle-income countries to service and reduce debts to commercial bank creditors. Available to low-income countries—the main objective was to restructure and buy back commercial debt with IDA credit. Considered extended repayment periods for lower-middle-income countries and rescheduling of ODA debt at a concessional rate. A proposal by the UK at the Commonwealth Finance Minister’s Meeting in Trinidad that would increase the grant element of debt reduction by up to 67 per cent from 20 per cent under the ‘Toronto terms’. Also known as ‘Enhanced Toronto’—it agreed to reduce debt service by 50 per cent on non-concessional bilateral debt with a 12-year grace period and 30 years maturity.

Paris Club—Toronto Terms, 1988 Brady Plan, 1989 IDA Debt Reduction Facility, 1989 Paris Club—Houston Terms, 1990 Trinidad Terms, 1990

London G7 Summit, 1991 Paris Club—Naples Terms, 1995

HIPC, 1996

Paris Club—Lyon Terms, 1996 Enhanced HIPC, 1999

Paris Club—Cologne Terms, 1999

A consensus about the UK proposal on Trinidad terms emerged—eligible countries would receive additional debt relief and debt service would be reduced by 67 per cent on non-concessional bilateral debt with a 16-year grace period and 40 years maturity. Debt stock reduction to bring the debt-to-export ratio under 200 per cent for 41 heavily indebted poor countries. The multilateral creditors for the ﬁrst time would take the initiative to forgo claims on their credits to help countries. Along with HIPC 1996, the Paris Club agreed to provide 80 per cent debt reduction in net present value terms. Increased stock reductions to bring the debt-to-export ratio down to below 150 per cent for a number of the 42 HIPC countries. Debt relief facility was made conditional on the formulation of comprehensive poverty reduction strategy papers (PRSPs). To support the Enhanced HIPC initiative, non-ODA credits were cancelled to a level of up to 90 per cent or more if necessary.

Source: Birdsall et al. (2002); Daseking and Powell (1999); Easterly (2002); and UNCTAD (2000).

hence called the Enhanced HIPC) in 1999 to provide faster and deeper debt relief to heavily indebted poor countries.11 Currently, there are 42 countries under the HIPC-II initiative (see Table 9.3); 27 of them have reached the ‘decision point’—when it is determined whether 11 The initiative is currently open to those poor countries that: (i) are eligible for highly concessional assistance such as from the World Bank’s International Development Association (IDA) and IMF’s Poverty Reduction Growth Facility (PRGF); (ii) face an unsustainable debt situation; and (iii) have a proven track record in implementing structural adjustment and poverty reduction strategies.

283

Mitigating the Impacts of Dependent Countries Table 9.3. Debt Relief for HIPC Countries

Countries

Total Reduction in Debt in NPV Terms (US$ million)

Nominal Debt Service Relief (US$ million)

Date of Completion Point/Decision Point

Countries that have reached their completion points (8) (as of 1 August 2003) Benin 265 460 Bolivia 1302 2060 Burkina Faso 553 930 Mali 539 895 Mauritania 622 1100 Mozambique 2023 4300 Tanzania 2026 3000 Uganda 1003 1950 Total for 8 countries

8333

14695

Countries that have reached decision points (19) Cameroon 1260 Chad 170 Congo, D.R.* 6300 Ethiopia 1275 Gambia 67 Ghana 2186 Guinea 545 Guinea-Bissau* 416 Guyana 585 Honduras 556 Madagascar 814 Malawi 643 Nicaragua 3267 Niger 521 Rwanda* 452 Sao Tome and Principe 97 Senegal 488 Sierra Leone* 600 Zambia 2499

2000 260 10000 1930 90 3700 800 790 590 900 1500 1000 4500 900 800 200 850 950 3850

Total for 19 countries

22741

35610

2224

3900

31074

50305

Countries still to be considered Burundi* ˆ te d’Ivoire Co Central African Republic Comoros Congo, Rep* Lao PDR Liberia* Myanmar* Somalia* Sudan* Togo Total Debt Relief Committed

Mar-03 Jun-01 Apr-02 Mar-03 Jun-02 Sep-01 Dec-01 May-00

Oct-00 May-01 Aug-03 Nov-01 Dec-00 Feb-02 Dec-02 Dec-00 Nov-00 Jul-00 Dec-00 Dec-00 Dec-00 Dec-00 Dec-00 Dec-00 Jun-00 Mar-02 Dec-00

Note: Four countries, Angola*, Kenya, Vietnam and Yemen, are considered as potentially sustainable cases without the HIPC assistance. * indicates conﬂict-affected countries. Source: Various World Bank and IMF Documents as available at www.worldbank.org/hipc

284

Commodity Prices and Debt Relief the level of debt is sustainable and, if unsustainable, how much debt relief will be required. Of the ‘decision point’ countries, eight have also attained the ‘completion point’—when countries receive their full package of debt relief beneﬁts.12 Under the Extended HIPC (HIPC-II) programme the external debt burden is deemed sustainable if, in general, the net present value (NPV) of debt does not exceed 150 per cent of exports.13 A highly open HIPC with a debt-toexport ratio of less than 150 per cent can qualify for the relief when its NPV of debt-to-revenue ratio reaches 250 per cent, with two further conditions: if it has (i) an export-GDP ratio of at least 30 per cent and (ii) a revenue-GDP ratio with a minimum threshold of 15 per cent.14 An important feature of HIPC-II is its conditionality, linking debt relief to policies for poverty alleviation. Countries beneﬁting from the scheme are required to establish a good record of implementing economic and social policy reforms and to prepare a poverty reduction strategy paper setting out ways and means of tackling the problem of poverty. PRSPs are supposed to envisage pro-poor complementary policies and compensatory expenditures.15 It is believed that savings out of the written-off outstanding stock and additional resource ﬂows will allow the beneﬁciary countries to devote more resources to social sectors such as health and education.16 As can be seen in Table 9.3, the currently eligible HIPCs are mostly from the continent of Africa—a total of 34. Of the remaining eight, Lao PDR, Vietnam, Myanmar and Yemen are Asian, and the rest—Bolivia, Guyana, Honduras and Nicaragua—are in Central and South America. Around three-quarters of these countries are LDCs (four—Comoros, Gambia, Guyana, and Sao Tome and Principe—are also small states) and countries classiﬁed as neither LDCs nor small states comprise about 20 per cent of the target sample (Bolivia, Cameroon, Republic of Congo, Ghana, Honduras, Kenya, Nicaragua, and Vietnam). 12 The HIPC initiative involves two stages—in the ﬁrst, the country builds a track record in implementing policies under the support and supervision of the World Bank and IMF for approximately three years and then it is decided whether or not its debts are sustainable. Based on performance of further policy reforms and structural adjustment, the relief package is delivered. In HIPC-I it was stipulated that countries would have to demonstrate their capacity for implementing reforms for another three years after the decision point. HIPC-II introduced the concept of the ‘ﬂoating completion point’, effectively making it possible to obtain debt relief earlier. Differences between the original and enhanced version of the HIPC schemes are discussed in detail in Gautam (2003) and UNCTAD (2000). 13 Since a signiﬁcant portion of the external debt stock is received on concessional terms, the NPV rather than the absolute amount is a better measure of the burden. The NPV is deﬁned as the sum of all future debt service obligations (interest and principal) on existing debt, discounted at the market interest rate. Since the interest rate is lower than the market rate, the NPV of debt for these poor countries is smaller than its face value. 14 Of the 27 decision point countries, assistance has been granted to 22 countries on the basis of the debt-to-export ratio condition. Five countries, Ghana, Guyana, Honduras, Mauritania, and Senegal, have been considered on the basis of the ﬁscal criteria (Dodhia, 2002). 15 Issues relating to pro-poor conditionality in HIPC may be found in Morrissey (2002). 16 According to the World Bank and IMF (2003), the decision point HIPCs are now spending approximately four times more on social services than on servicing debt.

285

Mitigating the Impacts of Dependent Countries The four HIPCs that are considered to be potentially viable without the debt relief package are Angola, Kenya, Vietnam, and Yemen. Table 9.3 shows that nominal debt reduction for 27 decision/completion point countries is estimated at about US$50 billion; in NPV terms the comparable ﬁgure is slightly over US$31 billion. The HIPC initiatives produce a marked improvement in the debt indicators of the recipient countries. After the relief operation, the mean NPV of debt-to-export ratio for 23 HIPCs for which information is available fell from a level much higher than that of the developing countries to a level at par with them (see Table 9.4). Similarly, the average NPV of debt-to-GDP ratio for the beneﬁciaries has fallen from as high as 60 per cent to 29 per cent. The burden of debt service in terms of export earnings has also declined substantially; the debt service-to-export ratio for HIPCs is estimated to be only 8 per cent in comparison with 21 per cent for the developing country group. The cost of debt relief under HIPC initiatives (up to March 2003) in 2002 NPV terms amounts to US$39.2 billion, which is almost equally divided between bilateral and multilateral creditors.17 Of the US$19.4 billion attributable to bilateral donors, US$14.5 billion is due to the Paris Club donors. On the other hand, the World Bank is to bear a cost of US$8.7 billion out of a total multilateral cost of US$18.8 billion; the IMF and the African Development Bank account for about US$3 billion each. In order to help multilateral lenders meet the cost, a special HIPC Trust Fund has been established with contributions from donor governments.18 Table 9.4. Debt Indicators in Developing Countries and HIPCs, 1999 (%) Decision point HIPCs (23)*

Indicators NPV of debt-to-exports ratio NPV of debt-to-GDP ratio Debt service-to-exports ratio

Developing country average

Non-HIPC developing countries

All HIPCs before HIPC relief

Before HIPC Relief (1999)

After HIPC Relief (2003)

133

128

249

259

127

38

36

84

60

29

20

21

14

16y

8*

Note: (1) Liberia and Somalia are always excluded from the group of HIPCs due to data unavailability; (2) *D R Congo, Ethiopia, Ghana and Sierra Leone are excluded; (3) y Average for 1998–99 based on debt service paid. Source: Abrego and Ross (2002). 17 The cost of the HIPC initiative is computed in NPV terms at the time of decision point. The cost is then increased every year after the decision point by a factor which is estimated as the average interest rate applicable for relief to be committed (IDA, 2003). 18 The Trust Fund will also comprise the lenders’ own resources. The IMF’s contribution to debt relief programmes will be funded in part from interest on the original subscriptions (in

286

Commodity Prices and Debt Relief

9.3. Commodity Prices and the HIPC Initiative The HIPC initiative has been subject to criticism ranging from the argument that ‘debt relief and other forms of aid are too great and too easy to get’ to that of ‘debt reduction being too small and tied to conditionality that is onerous and misguided’. Birdsall (2002) identiﬁes three other critiques relating to the structure under which the programme operates: (1) that the eligibility criteria of a country’s stock of debt as a share of its exports is inappropriate; (2) that as forgiving unpayable loans is an accepted reality, most of the debt reductions are to be treated as a loss (bad debt) rather than the relief provided; and (3) that debt sustainability analysis as carried out in the programme has been unrealistic. For the purpose of the present section of the report, criticisms related to (3) above deserve the most serious attention since, for reasons discussed below, this issue threatens to undermine the basic objective of the HIPC Initiative. As mentioned earlier, under the latest debt relief scheme emphasis is given to the issue of the sustainability of debt. Explicit attention has been given to determining the magnitude of debt stock needed to be written off to make any outstanding debt and new borrowing viable. The decision on the amount of relief that is thought to be sufﬁcient to achieve debt sustainability is made at the decision point and is updated at the completion point. While debt indicators such as NPV debt-to-exports and debt-to-revenue ratios are useful guidelines in assessing the severity of the burden for a country, reducing the ratios to a predetermined ‘sustainable’ level does not provide any automatic guarantee of debt sustainability in the medium to long run (UNCTAD, 2000). The medium-term prospective DSA is done on the basis of various assumptions about a number of key macroeconomic variables. These variables include real GDP growth, the income elasticity of demand for imports, growth in the volume and value of exports, and ﬂows of grants and foreign direct investment (FDI) and debt-creating ﬂows and conditionality attached to them. To attain sustainability it is required that the current account deﬁcit be covered by nondebt creating or concessional debt ﬂows so that the building up of debt stock can be avoided. Being projections, the DSA is sensitive to assumptions regarding the value of parameters. For example, UNCTAD (2000) shows that in the case of the United Republic of Tanzania, a small change in the export growth

gold) by members at the time of its inception. The balance will come from bilateral contributions. In the case of the World Bank, the earnings on its loans to middle-income countries will be used in the ﬁrst few years; after that it has to be supported by donor contributions. Of the others, because of lack of resources the African Development Bank will have to rely heavily on donors’ contributions; currently it can support only about 20 per cent of its contribution to the debt relief programme from its own resources.

287

Mitigating the Impacts of Dependent Countries

export growth (%)

200 10 150

5 0

100

50

250

Bolivia

12 200

10 8

150

6 100

4 2

50

Decision Point Projection: Export Growth

Updated Projection: Export Growth

Updated Projection: Export Growth

Decision Pt. Project.: NPV Debt/Exports (%)

Decision Pt. Project.: NPV Debt/Exports (%)

Updated Project.: NPV Debt/Exports (%)

Updated Project.: NPV Debt/Exports (%)

200

5 150 0 100

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

50 0

Mali

15

250

200

10 150 5 100 0 50

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

export growth (%)

250

10

20

300

export growth (%)

Burkina Faso

15

288

0

Decision Point Projection: Export Growth

NPV debt to export ratio (%)

20

0

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

0

NPV debt to export ratio (%)

15

14

0

Decision Point Projection: Export Growth

Decision Point Projection: Export Growth

Updated Projection: Export Growth

Updated Projection: Export Growth

Decision Pt. Project.: NPV Debt/Exports (%)

Decision Pt. Project.: NPV Debt/Exports (%)

Updated Project.: NPV Debt/Exports (%)

Updated Project.: NPV Debt/Exports (%)

NPV debt to export ratio (%)

250

export growth (%)

Benin

NPV debt to export ratio (%)

20

Commodity Prices and Debt Relief 350

200 5 150 0

100 50

80

30

60

20 10

40

0

20 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

100

40

0

Decision Point Projection: Export Growth

Updated Projection: Export Growth

Updated Projection: Export Growth

Decision Pt. Project.: NPV Debt/Exports (%)

Decision Pt. Project.: NPV Debt/Exports (%)

Updated Project.: NPV Debt/Exports (%)

Updated Project.: NPV Debt/Exports (%)

150

8 100

6 4

50

2008 2009 2010

2005 2006 2007

2003 2004

2 0

Uganda

15

export growth (%)

10

NPV debt to export ratio (%)

200

12

2001 2002

20

250

Tanzania

2000

120

50

0

14

export growth (%)

140

60

Decision Point Projection: Export Growth

16

0

160

70

NPV debt to export ratio (%)

10

180

80

300 250

10 200

5

150

0

100 50

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

export growth (%)

250

90

NPV debt to export ratio (%)

300

15

Mozambique

export growth (%)

Mauritania

NPV debt to export ratio (%)

20

0

Decision Point Projection: Export Growth

Decision Point Projection: Export Growth

Updated Projection: Export Growth

Updated Projection: Export Growth

Decision Pt. Project.: NPV Debt/Exports (%)

Decision Pt. Project.: NPV Debt/Exports (%)

Updated Project.: NPV Debt/Exports (%)

Updated Project.: NPV Debt/Exports (%)

Figure 9.2. Projected Export Growth and NPV Debt-to-Export Ratio for Countries that have Reached Completion Point Note : The NPV debt-to-export ratio is measured on the right vertical axis while the export growth rates are measured on the left vertical axis. Source : Data on decision point and updated projections are taken from IMF and IDA (2002).

289

Mitigating the Impacts of Dependent Countries target could have substantial effects.19 In fact, the assumptions that are crucial to the outcomes are the level and terms of new ﬁnancing and an assessment of future economic and export performance. Already, there is some evidence that assumptions with regard to future growth rates of exports and GDP may have been too optimistic. Figure 9.2 gives the decision point projected export growth rate and net debt-to-exportratios vis-a`-vis the comparable updated projections at the completion point for eight HIPC countries. As can be seen, the updated ﬁgures for export growth for 2000–03, which more closely reﬂect reality, differ quite signiﬁcantly from the original projections. For Benin, the export growth rate for 2000 was projected to be about 5 per cent, but actually proved to be 12.5 per cent; Bolivia’s 10 per cent export growth projection in 2001 compares with an actual performance of 0.9 per cent. Similar contrasts between projected and updated ﬁgures also appear for Mauritania (2002), Mozambique (2002), Tanzania (2000, 2002 and 2003) and Uganda (2000, 2001 and 2002). These ﬁgures suggest that the decision point assumptions do not even match the actual performance in the near future, making the basis for projections for the distant and long-time periods (e.g. 2005–2017) less than informative. The unweighted average projected export growth rate during 2000–2010 for a number of 24 HIPCs is envisaged to be 8.6 which is just about the half the rate achieved by these countries in the decade of the 1990s (see Figure 9.3). Figure 9.3 shows that in order to achieve the projected growth, given their recent performance, HIPCs need to demonstrate unusually robust performance in the next decade. The unweighted average export growth for the 24 countries as shown in Figure 9.3 in 2000–01, estimated at 5.4 per cent, appears to be signiﬁcantly lower than the 9.4 per cent projected at the decision point (IMF and IDA, 2002). Lower than projected exports in 2001 were observed for 16 out of 24 countries, while only six recorded a better than projected export performance. According to estimates by the IMF and World Bank, the shortfalls in export revenue were particularly large for Uganda (27 per cent), Guinea, Senegal and Zambia (16–18 per cent), Burkina Faso and Guinea Bissau (20 per cent), Benin, Honduras and Mauritania (10–13 per cent). Lower exports for these countries reduced the basis for export projections in the medium term, thus shifting downwards the level of projected exports and, with other factors remaining constant, moving the projected NPV of debt-to-exports ratios upwards. An evaluation by the IDA and the IMF, therefore, observes that the recent global economic slowdown, coupled with a signiﬁcant decline in many primary commodity prices, has weakened the HIPCs’ growth and export performance 19 In the UNCTAD sensitivity analysis, the projected growth of exports of 8.3 per cent per annum for 2000–2018 was reduced by only 10 per cent to generate a ﬁnancial gap 120 per cent higher than the baseline forecast.

290

Commodity Prices and Debt Relief

Weighted average

Zambia

Simple average

Tanzania Senegal

Rwanda

Nicaragua Niger

Mozambique Mauritania

Malawi Mali

Honduras

Madagascar

Guinea-Bissau Guinea

Burkina Faso

Gambia

Guyana

5

Cameroon

per cent

10

Benin Bolivia

15

Uganda

Projected for 2000–10 Sao Tome and Principe

1990–99 actual average 20

0

−5

Figure 9.3. Average 1990–99 Actual Export Growth Rate vis-a`-vis Projected Growth Rate for HIPCs Note : Actual growth rates are based on nominal US dollars for exports of goods and non-factor services using the World Bank data. Projected growth rates are from World-IMF documents on HIPCs.

in the last two years and led to a deterioration of the external debt indicators for many of these countries (IMF and IDA, 2002). The importance of export performance to debt sustainability of HIPCs is best illustrated by the experience of Uganda. Uganda was the ﬁrst country to qualify for debt relief under HIPC-I and it also reached completion point under HIPC-II before any other country. In 2000 Uganda’s export growth rate was approximately 14 per cent and this was followed by a further decline of 4 per cent in 2001, as against a strong recovery projection of about 15 per cent. The updated growth for 2002 is two and a half times lower than the decision point projection. As a result, in 2002 Uganda’s NPV of debt-to-export rose to 254 per cent—considerably more than the debt sustainability target of 150 per cent—and, more detrimentally, much higher than the decision point projected ratio of 97 per cent. This weak performance cannot be attributed to poor economic management. Rather, it was due mainly to the secular decline in coffee prices, which resulted in the loss of signiﬁcant export earnings to Uganda (Birdsall, 2002). It follows from the above that exogenous external shocks resulting from a secular fall, together with frequently volatile movements in commodity

291

Mitigating the Impacts of Dependent Countries prices, are likely to remain as the most important threat to countries maintaining a sustainable debt target in the future. While differences in the evolution of the debt indicators among HIPCs might reﬂect differences in the implementation of various reform programmes, the external debt sustainability outlook for most of these countries critically hinges on their export structure and performance. As most of these countries have a degree of dependence on commodities for exports, they are most likely to be subject to adverse commodity price shocks. Downward trends in the prices of primary commodities would also imply that in the absence of diversiﬁcation in the composition of their exports, these countries will continue to be marginalized in world export trade.20 This is further accentuated by the fact that due to low income elasticity, the share of primary commodities in world merchandise is in secular decline.21 In fact, it appears that one of the most important omissions in the formulation of the HIPC Initiative has been the lack of recognition of the problem posed by trends in commodity prices and of a policy framework to address it. Although the linking of debt relief to poverty reduction is desired, the problem of debt sustainability is ultimately tied to a country’s potential foreign exchange earning capacity. The policy conditionality of HIPC–PRSP, however, has only an indirect inﬂuence by aiming to provide a stable macroeconomic environment. Although the principal objective of growth in a low-income country is poverty alleviation, in the short run the link between poverty reduction efforts and debt sustainability is obscure, while the impact of export earnings on debt indicators is immediate. In an extreme case, a country can borrow to invest in the non-tradable sector only to alleviate poverty effectively, but future debt servicing will be difﬁcult if exports do not expand, even though the poverty situation may improve.22 Allowing for sufﬁcient time, poverty alleviation and economic growth can energize exports, but the problem is that if the debt is not serviced immediately and continuously a country may well slip back into the trap of new borrowing to service the accumulated debt. This is all the more likely, because despite the HIPC relief, the absolute amount of debt service payment for HIPCs in the medium term will be higher than in the pre-HIPC period. Table 9.5 compares the information on actual and projected debt service in absolute terms, as well as in terms of proportion of exports and GDP. It is observed that for Chad, Niger, Sierra Leone, and Zambia, debt service payments in 2005 will be higher than the corresponding actual payments made in 1998. For Burkina Faso, Cameroon, 20 Grynberg and Razzaque (2003) provide evidence of statistical association between LDCs’ and small states’ marginalization in world export trade and the diminishing share of agricultural commodities in world merchandise trade. 21 In 1980 agricultural products constituted about 16 per cent of world merchandise exports, whereas they now account for about 7 per cent. 22 It has also been argued that the debt relief initiative may not even be able to meet the poverty reduction targets (Serieux, 2001). In the context of Kenya, Kiringai (2002) argues that the process of the preparation of a PRSP only enhances the stakeholders’ expectations, but cannot deliver.

292

Commodity Prices and Debt Relief Honduras, Mali, Rwanda, Senegal, Tanzania and Uganda, these payments will comprise substantial proportions of their actual payments in the pre-HIPC period. If overall exports and resource ﬂow grow as projected, the debt indicators will not be worrying, but how far export growth targets can be maintained remains a big question. For commodity-dependent countries, the role of trade policy in development should be well integrated into its poverty reduction strategy. Understandably, most HIPCs have a natural static comparative advantage in the primary sector, and in the past many pursued inward-looking import-substitution strategies to develop a domestic industrial base. The import-substituting industrialization strategy developed under the protective regime remained inefﬁcient, and in the face of severe external and internal imbalances a policy of trade liberalization and reform was carried out. Since the import-substitution regime resulted in policy-induced biases against agriculture, a policy reversal to a strategy of export promotion served to revive and enhance the static comparative advantage of primary commodities. The export structure remained undiversiﬁed and, in most cases, has not allowed poor countries to take full advantage of high income growth in the world economy. The challenge for the HIPCs is to formulate an effective strategy that enables them to reduce dependence on commodities through diversiﬁcation with the help of policy and incentive mechanisms in a market-friendly way and not in the way of traditional import-substitution. Do PRSPs address the issue of trade policy and diversiﬁcation? Reviewing the strategies for a number of countries, Hewitt and Gillson (2003) ﬁnd that the PRSPs designed by different countries do not give adequate attention to trade policy. In general, there has been some emphasis on the promotion of exports, which may not have any signiﬁcant effect on the objective of diversiﬁcation. For countries in the interim period, the enhanced HIPC initiative allows some ﬂexibility in exceptional cases to top-up debt relief at the completion point where exogenous factors have caused fundamental changes in their economic circumstances (known as ‘topping-up’) (IMF and IDA, 2002).23 But it does not contain any provision for addressing this problem beyond completion point. It is rather difﬁcult to conceive that the effect of exogenous shocks can be mitigated in the post-completion stage. The bottom line of our argument is: while there is no doubt that linking debt relief with enhanced poverty reduction efforts and increased social expenditure is commendable, the importance of foreign exchange earnings, which is fundamental to the countries’ long-term external debt sustainability, has not been sufﬁciently emphasized.24 23 Under this provision Burkina Faso was granted an additional US$128 million at completion point. 24 As Ranis and Stewart (2001: 2) also stress, ‘ . . . in a sense it [the HIPC Initiative] is answering the wrong question—the problem is not debt, nor insufﬁcient ﬂows of externally unearned resources, but lies elsewhere, in the failure of countries to earn foreign exchange, and in deep structural and political problems that are not addressed by HIPC’. Similarly, Geda

293

Table 9.5. Actual and Projected Debt Service Indicators for HIPCs that have Reached Decision Point Debt Service (in US$million) Actual

Debt service/exports (in per cent)

Projection

Actual

Debt service/GDP (in per cent)

Projection

Actual

Projection

Countries

1998

2002

2003

2005

1998

2002

2003

2005

1998

2002

2003

2005

Benin Bolivia Burkina Faso Cameroon Chad Ethiopia Gambia Ghana Guinea Guinea-Bissau Guyana Honduras Madagascar Malawi Mali Mauritania Mozambique Nicaragua Niger Rwanda Sao Tome and Principe Senegal Sierra Leone Tanzania Uganda Zambia

64.1 390 60 401 26 101 26.1 560.1 128.2 7 130.8 311.2 166.1 90.1 74 88 104 231.4 17 18 6.6 207 8.9 224 110 147.3

33 270 22.5 228.9 25.8 149 16.1 267 85.6 2.2 59 220.5 50.5 47.2 69.5 39.2 40.2 158 25.4 13 2 141 19.7 106.9 60 137.7

30.9 290 19.1 225.2 32.7 88 16.1 164.2 81.6 5.3 46 315.2 53.6 79.4 63.4 35.1 47.3 11 26 13.7 2.1 146.4 27.6 157.7 64.3 177.8

33.5 277 25.1 253.6 36.4 88 12 111.6 68.4 3.3 37 202.4 72.7 50.7 64.9 36.5 61.9 8.3 29 15.7 1 138.9 13.5 166.3 78.1 215.1

16.1 28.6 16.5 18 8 9.7 12.4 22.1 15.5 23.5 19 12.6 20.6 15.6 11.5 22 41 8.2 5 30 55 15 9.4 20.7 15 16

10 16.1 7.4 9.2 11 15.1 12.5 10.2 10.8 3.7 8.8 9.1 6.6 9.7 6.7 10.6 3.8 16.9 7.2 9.6 8.7 9.2 14.1 6.9 8.6 12.7

6.8 17.3 5.2 7.7 12 9.2 11.9 5.6 9.4 8.1 6.8 13.3 5.2 15 6.2 9.1 4.3 11 6.7 9.1 8.6 8.8 18.2 8.8 8 14.2

6 14.1 5.9 8.9 2.1 9.3 8.2 3.4 7 3.8 5.5 8.3 5.9 8.6 5.3 8.1 2.6 8.3 6.8 9 3.4 7.3 5.7 7.8 8.2 13.5

2.8 4.5 2.3 4 2.2 1.5 6.2 7.5 3.6 3.4 18.2 5.9 4.3 5.1 2.8 10 2.5 10.5 1.3 2 16.3 4 1.3 2.8 1.7 4.5

1.2 3.5 0.8 2.5 1.3 2.4 5.8 4.3 2.7 1 8.8 3.3 1.1 2.5 2.2 3.9 1 6.1 1.2 0.8 3.7 2.8 2.5 1.2 1 3.7

1 3.8 0.6 2.1 1.4 1.3 6.1 2.4 2.5 2.1 6.2 4.5 1.1 4.5 1.8 3.2 1.1 4.3 1.1 0.8 3.4 2.5 3 1.7 1.1 4.8

0.9 3.5 0.6 2 0.9 1.2 4 1.3 1.8 1.1 4.7 2.5 1.3 2.5 1.6 3 1.2 3.5 1 0.8 1.4 2 1.3 1.6 1.2 5.2

2250.5

2296.6

2169.1

17.5

9.9

9.2

7

4.1

2.4

2.3

1.9

All (26 Countries)

3673

Note: For all 26 countries the debt service-export and debt service-GDP ratios are a weighted average. Source: Authors’ compilation from various World Bank and IMF documents available at www.worldbank.org/hipc

Commodity Prices and Debt Relief

9.4. Incorporating a Real Price Adjustment Mechanism in the HIPC Initiative to Address the Problem of Weakness in Commodity Prices It follows from the above discussion that the debt relief programmes for primary producing HIPC members can only be effective if they are complemented by external support in the event of external shocks. While unfavourable production shocks cannot be predicted far in advance, in the light of historical experience, it is very likely that a secular downward trend in prices, together with sharp price ﬂuctuations, will tend to persist. Therefore, it is important to have a support mechanism for the HIPCs to compensate for an adverse trend in commodity prices. Gilbert and Tabova (2003) outline a number of modalities for linking commodity prices with countries’ ability to service debt using various ﬁnancial derivatives.25 These modalities appear to be complex and may require further motivations and reﬁnements. The nature of the empirical operation of the schemes also appears to be complicated. Despite the attractive features of Gilbert and Tabova’s modalities, the authors’ simulated results provide only mixed results and the authors themselves have refrained from making any recommendation on the adoption of such a scheme. The area of commodities is one where many different schemes have been tried in the past, but they have all virtually collapsed without providing any clear indication about corrective measures which would have made them viable. Excessive distortionary measures in commodity markets may also be untenable, especially as under the current WTO trade regime there is enormous pressure for liberalization of trade in agriculture. Rather than emphasizing any new initiatives, it appears more feasible to consider the impact of commodity prices on the debt burden of the predominantly commodity-exporting countries under the existing HIPC Initiative. Birdsall, Williamson and Deese (2002) propose such a provision, calling for additional debt relief for decision-point HIPCs in the face of adverse exogenous shocks.26 It is suggested that if in any year any of the HIPCs’ (2002) concludes that the African debt problem is essentially a trade problem and consequently the long-run solution to debt points to the importance of addressing trade and traderelated structural problems. 25 Gilbert and Tabova (2003) considered two schemes: commodity swaps and commodity swaptions. Commodity swaps are similar to loans taken by mining companies where both borrowing and repayment take place in terms of gold. There can be a gold interest and since transactions are made in terms of the quantity of gold, the arrangement combines ﬁnancing and hedging. A gold loan is to be considered as a normal currency loan plus a ﬁxed-for-ﬂoating gold swap. In the case of countries dependent on other commodities for exports, their debt service is independent of commodity prices and hence the countries might beneﬁt from swapping out this ﬁxed rate exposure for a ﬂoating rate which matches their ﬂoating commodity price exposure. Commodity swaptions, on the other hand, are schemes with the principle of coping with exceptional price movements only. 26 Birdsall et al. (2002) also propose that other exogenous shocks, such as weather, should be considered.

295

Mitigating the Impacts of Dependent Countries debt service to GNP ratio exceeds 2 per cent, the World Bank/IMF would examine whether the unfavourable trend could be attributable to exogenous factors.27 In the case of a positive assessment, compensation might be provided to the countries to reduce their debts to a sustainable level. Modifying the Birdsall, Williamson, and Deese (2002) proposal, we propose that there should at least be a compensatory mechanism for providing additional debt relief to decision point HIPCs when they are subjected to export revenue shortfall due to adverse trends in commodity prices. Establishment of such a support mechanism is justiﬁed given the secular decline and volatile nature of commodity prices and the resulting loss of purchasing power of exports, as well as earnings instability, of the commodity-dependent poor countries. The proposed extension to the HIPC Initiative is found to be relatively low cost, as estimated in the next section, providing further justiﬁcation for this additional support to beneﬁciaries.

9.4.1. Hypothetical cost of a real commodity price adjustment mechanism under the HIPC initiative Ideally, using the trend rate of decline in commodity prices and some quantity of output considered to be the export of a normal year, one can estimate the effect of weakness in commodity prices and work out the compensation needed by individual decision point countries to keep their debt service at a sustainable level. However, trend rates are sensitive to the choice of sample period and are frequently accompanied by a much sharper fall in price than the average rate.28 Estimation based on the trend rate will also require the consideration of several export commodity prices for each of the individual countries, making it an extremely data intensive and complicated procedure. Furthermore, changes in export quantity might allow a country to maintain a sustainable debt burden for some period despite weakness in prices. Taking all these factors into account, we calculate the cost of a compensatory debt relief mechanism for HIPCs based on their export earnings from commodities.

27 The ‘2 per cent of GNP’ threshold comes from the argument of inappropriateness of the NPV of debt-to-export ratio as an adequate indicator of a country’s debt-servicing burden. It is suggested that a country should not be expected to spend more than 10 per cent of government revenue on debt service. Since most HIPCs in 1999 collected about 20 per cent of their GNP in tax revenue, the sustainable debt-servicing ratio can then be calculated as 2 per cent of GNP. 28 Therefore, a compensation based on the long-term trend rate can result in lower debt relief for a country in periods with much bigger adverse shocks in prices than the one suggested by the trend.

296

Table 9.6. Hypothetical Cost of Compensation for HIPCs Share of Primary Export Projection for HIPCs (US$ million) Exports Primary in Total Exports Exports Actual export Avg: 1996– in 2001 growth rate: (US$ 2000 million) 2001 2010 1990–99 (%)b (ratio)c a

Countries (1) Benin Bolivia Burkina Faso Cameroon Chad Ethiopia Gambia Ghana Guinea Guinea-Bissau Guyana Honduras Madagascar Malawi Mali Mauritania Mozambique Nicaragua Niger Rwanda Sao Tome and Principe Senegal Sierra Leone Tanzania Uganda Zambia

Primary Exports in 2010 if Grown at a Rate of 1990–99 (US$ million)

Projected Primary Exports in 2010 (US$ million)d

Shortfall in Primary Exports in 2010e

Revised Total 2010 Exports Debt in 2010f Stockg

Revised Debt-toExport Ratio in 2010

Total Exports Required for Debt-toExport Compensatory Ratio to be Debt Relief 150 per centh (US$ million)i

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

(13)

392 1442 305 2586 242 952 128 2416 860 71 718 2673 1046 480 662 433 805 932 279 126 18

791 3108 751 4248 1978 1815 233 4597 1647 181 1037 5456 1811 763 1190 528 3451 1570 484 367 42

2.5 3.6 2.6 0.0 0.6 2.6 2.7 11.1 1.0 7.5 5.0 8.5 6.5 2.1 2.3 2.5 6.8 10.0 4.5 3.0 5.0

0.4 0.7 0.6 0.6 0.6 0.9 0.9 0.6 0.6 0.7 0.7 0.5 0.8 0.8 0.5 0.6 0.6 0.8 0.7 0.8 0.8

156.8 980.6 180.0 1474.0 133.1 866.3 115.2 1425.4 481.6 51.1 473.9 1283.0 868.2 384.0 357.5 251.1 458.9 708.3 203.7 97.0 13.5

191.8 1310.1 146.6 1482.4 140.4 1073.5 143.8 3328.3 446.9 91.8 704.8 2476.0 1448.8 457.1 431.1 206.3 781.1 1529.9 141.9 76.6 20.1

316.4 2113.4 443.1 2421.4 1087.9 1651.7 209.7 2712.2 922.3 130.3 684.4 2618.9 1503.1 610.4 642.6 306.2 1967.1 1193.2 353.3 282.6 31.5

124.6 803.4 296.5 938.9 947.5 578.2 65.9 616.0 475.4 38.5 20.3 142.8 54.4 153.3 211.5 100.0 1186.0 336.7 211.4 206.0 11.4

666.4 2304.6 454.5 3309.1 1030.5 1236.8 167.1 5213.0 1171.6 142.5 1057.3 5313.2 1756.6 609.7 978.5 428.0 2265.0 1906.7 272.6 161.0 30.6

795.0 3333.0 1024.0 4248.0 934.0 2439.0 301.0 3503.0 1565.0 248.0 736.0 3323.0 1929.0 1148.0 1520.0 656.0 1611.0 1712.0 768.0 541.0 59.0

1.2 1.4 2.3 1.3 0.9 2.0 1.8 0.7 1.3 1.7 0.7 0.6 1.1 1.9 1.6 1.5 0.7 0.9 2.8 3.4 1.9

530.0 2222.0 682.7 2832.0 622.7 1626.0 200.7 2335.3 1043.3 165.3 490.7 2215.3 1286.0 765.3 1013.3 437.3 1074.0 1141.3 512.0 360.7 39.3

1692 121 1194 801 1038

2765 330 2274 1953 2207

1.0 5.0 7.9 11.5 3.0

0.5 0.4 0.7 0.9 0.7

778.3 49.6 871.6 688.9 737.0

721.5 33.0 1613.1 1659.8 581.7

1271.9 135.3 1660.0 1679.6 1567.0

550.4 102.3 46.9 19.8 985.3

2214.6 227.7 2227.1 1933.2 1221.7

2364.0 127.0 3525.0 1320.0 2575.0

1.1 0.6 1.6 0.7 2.1

1576.0 84.7 2350.0 880.0 1716.7

Total Compensatory Debt Relief (2001–2010)

(14)

228.1

389.2 33.5

22.8

155.6 34.8 9.3

239.4 199.6 8.7

122.9 494.9 1939.0

Note: a Projections made during decision point; b exports including both primaries and fuels; c estimated from UNCTAD database; d based on the assumption that share of primary exports in total exports remains unchanged; e is the difference between columns (8) and (7); f (3) less (10); g projected at the decision point; h such that ﬁgures in column 13 divided by 11 come out to be 150 per cent; i compensation is estimated for only those countries with revised debt-to-export ratio higher than 150 per cent.

Mitigating the Impacts of Dependent Countries Table 9.6 gives the details of the hypothetical cost estimation procedure. Following Birdsall, Williamson and Deese (2002), the present proposal considers the operation of a real commodity price adjustment mechanism for a period spanning over ten years, i.e. 2001–10.29 Column 6 of Table 9.6 gives the export earnings from primary commodities for individual decision point HIPCs in 2001. Since the DSA export growth projections have been too optimistic, as argued above, we estimate export earnings from primary commodities for each of the HIPCs in 2010 based on their actual export growth rate during 1990–99 (column 7). The assumption of unchanging share of commodities in total exports allows the computation of the value of primary exports in 2010 (column 8) based on the World Bank/IMF projection for total export earnings (column 3). The shortfall (column 9) in primary exports is then estimated as the difference between the projections based on DSA growth rate and those based on actual growth rate in the 1990s. The shortfall in export earnings is deducted from the projected exports of 2010 to obtain the ﬁgure for revised total exports shown in column (10).30 With the help of the projected total debt stock (column 9), it is now possible to obtain the debt-to-export ratio (column 12). Column (13) then calculates the export earnings required to make the debt burden just sustainable, which is then compared with the revised total exports in column (10) to obtain the adjustment to debt relief. Naturally, the adjustment is to be made only for those countries with a debt-toexport ratio higher than 150 per cent. The total solely due to deteriorating commodity export earnings for the 26 decision point countries involving a time horizon of 2001–10 is estimated at approximately US$2 billion, i.e. about US$0.2 billion a year.31

9.5. Conclusion While poverty alleviation should be the prime target of growth and development in poor countries, the role of trade and particularly enhanced export earnings cannot be overemphasized if any debt relief effort is to be made sustainable in the long run. The current HIPC scheme does not consider how the objective of rising exports will be achieved, given the nature of HIPC dependence on commodity exports and the past trends associated with their prices which has resulted in instability of export earnings and declining real export prices.

29

The extension over any time period is quite straightforward. This, in essence, assumes that non-primary exports grow at a rate as depicted in the DSA. The comparable computation by Birdsall, Williamson and Deese (2002), who consider shocks in total exports rather than the earnings from primary exports only, appears to be US$5.2 billion. 30 31

298

Commodity Prices and Debt Relief Currently, the HIPC scheme does not consider the possibility of postcompletion point beneﬁciary countries’ exposure to terms of trade shocks due to commodity price collapse and the resultant unsustainable debt burden. It is, however, realistic and feasible to consider an adjustment mechanism for adverse shocks in commodity export earnings so that the HIPC graduates can receive additional debt relief in cases where falling prices result in unsustainable debt. Finally, two other issues are of utmost importance. First, given the problems associated with commodities and the excessive dependence of LDCs and small states on primary commodities, the permanent solution to the problem of debt crisis lies in a structural shift in the composition of the export basket of these countries. Second, the debt relief programme should not be considered as a replacement for donor aid ﬂows. Without additional aid, the enhanced ﬁscal capacity needed for social and other expenditures may not be achieved.

Appendix 9.1. Debt Indicators for 11 HIPCs that are still to be considered in the Debt Relief Programme COUNTRY

NPV of Debt stock 2001 (US$ Million)

PV/GNI

PV/XGS

EDT/GNI

1065 10647 536 177 4232 1295 1928 4032 2531 14547 999 41989

95 102 55 82 218 82 472 55 n.a. 148 79 —

1090 220 499 294 182 263 1679 174 n.a 673 205 —

156 111 84 82 231 157 487 78 n.a. 156 111 —

Burundi ˆ te d’Ivoire Co Central African Rep. Comoros Congo Rep. Lao PDR Liberia Myanmar Somalia Sudan Togo Total Debt

Note: PV stands for present value. GNI is gross national income; XGS is exports of goods and services; EDT is total external debt including short-term and the use of IMF credit. Source: Global Development Finance, 2003.

299

Mitigating the Impacts of Dependent Countries Appendix 9.2. Debt Indicators for non-HIPC LDCs COUNTRY

NPV of Debt Stock 2001 (US$ million)

PV/XGS

EDT/GNI

PV/GNI

n.a. 9712 245 2301 235 817 1567 n.a. 231 177 193 n.a. 406 177 142 112 37

n.a 113 151 159 82 154 92 n.a. 97 72 8 n.a. 74 38 148 68 21

n.a 33 53 85 61 32 49 n.a 63 46 54 n.a 55 43 85 58 31

n.a 21 49 72 35 21 28 n.a 41 31 44 n.a 37 32 59 40 17

Afghanistan Bangladesh Bhutan Cambodia Eritrea Haiti Nepal Tuvalu Cape Verde Djibouti Equatorial Guinea Kiribati Lesotho Maldives Samoa Solomon Islands Vanuatu Total

16352

Note : PV stands for present value. GNI is gross national income; XGS is exports of goods and services; EDT is total external debt including short-term and the use of IMF credit. Source: Global Development Finance, 2003.

Appendix 9.3. Debt Indicators for non-HIPC Small States COUNTRY

NPV of Debt Stock 2001 (US$ million)

PV/XGS

EDT/GNI

PV/GNI

n.a 739 765 307 231 177 181 193 175 190 5361 n.a. 406 177 1357 1658 2188 142 212 112 n.a. 297 42 2609 37

n.a. 49 191 9 97 72 125 8 15 79 120 n.a. 74 38 32 60 98 148 45 68 n.a. 27 45 59 21

n.a. 29 102 8 63 46 87 54 11 59 68 n.a 55 43 43 40 80 85 37 58 n.a. 22 42 33 31

n.a. 30 110 6 41 31 76 44 10 52 73 n.a 37 32 38 38 70 59 37 40 n.a. 22 28 35 17

Antigua and Barbuda Barbados Belize Botswana Cape Verde Djibouti Dominica Equatorial Guinea Fiji Grenada Jamaica Kiribati Lesotho Maldives Malta Mauritius PNG Samoa Seychelles Solomon Islands Suriname Swaziland Tonga Trinidad and Tobago Vanuatu

Note: PV stands for present value. GNI is gross national income; XGS is exports of goods and services; EDT is total external debt including short-term and the use of IMF credit. Source: Global Development Finance, 2003.

300

10 Aid Flows and Commodity Prices Mohammad A. Razzaque, Philip Osafa-Kwaako, and Roman Grynberg

The persistent weakness of real commodity prices presents serious challenges for export earnings and domestic incomes in predominantly commodity exporting developing countries, as indicated by the estimates provided in Chapter 6 of substantial foreign exchange losses for these countries. In most of these countries, the budgetary position of national governments or monetary authorities is highly sensitive to their primary commodity export earnings and aid ﬂows.1 In principle, the effects of declining commodity prices may be offset by increased aid ﬂows, and following the effective collapse of most international commodity agreements and the end of compensatory ﬁnance arrangements (such as STABEX), there has been renewed interest in developing alternative concessionary aid-supported schemes to compensate for some of the terms of trade losses incurred by commodity-dependent countries.2 In this chapter, issues related to aid ﬂows and commodity prices are considered from two perspectives. First, the pattern of recent aid ﬂows to developing countries that are heavily dependent on primary commodities is examined to assess whether the secular decline in commodity prices has been accompanied by any commensurate increases in aid ﬂows. Second, we propose the establishment of a multilateral aid instrument that would provide additional funds to commodity-dependent countries. In contrast to the risk management tools designed to address the problem of price volatility, the prolonged and secular decline in real commodity prices requires policies aimed at long-term structural diversiﬁcation. Our present proposal is not to create an export earnings stabilization fund, but to provide additional 1 This is partly the result of the low tax revenues obtained in most developing countries (Agenor and Montiel, 1999). 2 Proposals for the establishment of new STABEX-like compensatory arrangements have been suggested under the ‘Chirac Initiative’, and in a report to the European Parliament Research Directorate prepared by the Overseas Development Institute (Page and Hewitt, 2001).

301

Mitigating the Impacts of Dependent Countries resources for diversiﬁcation projects in the low-income and highly vulnerable developing countries that rely predominantly on commodities for domestic production and export. The design of such a scheme may be envisioned in a number of ways, with different burden-sharing arrangements among donors, and different implications for recipient countries. What is proposed is one option for addressing the many issues involved in establishing such a scheme and is not intended to be deﬁnitive, but rather a basis for subsequent debate. The underlying principles in the proposal are outlined, together with some preliminary assessment of the cost of putting such a scheme into operation.

10.1. Aid Flows and Declining Commodity Prices For most developing countries, aid ﬂows, in real terms, fell in the 1990s (White, 2002).3 Figures 10.1–10.3 show the relationship between the composite relative commodity price index and aid ﬂows to LDCs, HIPCs, and small states respectively. The left vertical axis in these graphs measures the aid ﬂows in real terms (in 1985 prices), while the right vertical axis is the scale of measurement for the composite relative commodity price index. A cursory assessment of the plots indicates that declining commodity prices have not been compensated for by commensurate increases in aid ﬂows. Although real aid ﬂows to HIPCs and LDCs showed modest growth in the 1980s, in the 1990s they fell, thereby compounding the economic impact of the simultaneous weakness in relative commodity prices. The pattern of declining aid ﬂows is most prominent in the case of small vulnerable states (Figure 10.3). Apart from a temporary increase in the early 1990s, aid ﬂows to SVs have experienced a sustained decline for most of the period since 1980. As an analysis based purely upon a composite index of export prices might miss individual price trends in the group of LDCs, HIPCs and SVs, we examine, for expository purposes, three country-speciﬁc cases. Figures 10.4–10.6 give a graphical exposition of aid ﬂows to Mali, Papua New Guinea and Togo, and the real prices of signiﬁcant national export commodities, viz. cotton, cocoa and phosphate respectively.4 In the case of Mali, cotton contributed an annual 3 We use the OECD deﬁnition of aid as ‘ofﬁcial development assistance’, referring to the sum of grants, concessional loans (with a grant element of more than 25 per cent), food aid and technical cooperation (Cassen et al., 1994). Other measures of aid such as ‘effective development assistance’ (EDA) and ‘ofﬁcial development ﬁnance’ (ODF) exist in the development ﬁnance literature. The World Bank’s most recent measure of EDA refers to the sum of grants and the grant element present in concessional loans disbursed within a given time period (Chang et al., 1999). On the other hand, ODF refers to the sum of EDA and the non-grant component of ofﬁcial loans. Compared with EDA, both ODA and ODF tend to overestimate aid ﬂows to recipient countries as the face value of concessional loans is included in these measures. 4 A more rigorous exercise would require the construction of country-speciﬁc aggregate relative price indices reﬂecting individual countries’ export baskets.

302

Aid Flows and Commodity Prices 12000

1.4

10000

US$ Millions

1.2 8000

1

6000

0.8 0.6

4000

0.4 2000

0.2 LDC Aid Flows

Composite Commodity Price Index

0 1980

1985

Relative Commodity Price Index

1.6

1990

1995

0 2000

Figure 10.1. Composite Relative Commodity Prices and Aid Flows to LDCs, 1980-2000.

14000

1.6

12000

1.4 1.2

US$ Millions

10000

1 8000 0.8 6000 0.6 4000

0.4

2000

0.2 HIPC Aid Flows

Relative Commodity Price Index

Note : The left vertical axis measures aid ﬂows, while the right vertical axis measures the aggregate relative price index. Aid ﬂows are in terms of 1985 real prices. Source : Real aid ﬂows are authors’ estimates based on data from UNCTAD (2002b), while the relative commodity price index is as constructed in Chapter 3.

Composite Commodity Price Index

0

0 1980

1985

1990

1995

2000

Figure 10.2. Composite Relative Commodity Prices and Aid Flows to HIPCs, 1980–2000. Note : The left vertical axis measures aid ﬂows, while the right vertical axis measures the aggregate relative price index. Aid ﬂows are in terms of 1985 real prices. Source : Real aid ﬂows are authors’ estimates based on data from UNCTAD (2002b), while the relative commodity price index is as constructed in Chapter 3.

average of 62 per cent of national export earnings over the period 1995–2000. Real cotton prices have, however, declined markedly since 1980, resulting in substantial foreign exchange losses for the country.5 Aid ﬂows to Mali, 5 The trend decline rate in the real cotton price is estimated in Chapter 3, while estimates of foreign exchange losses are provided in Chapter 6.

303

1800

1.6

1600

1.4

US$ Millions

1400

1.2

1200

1

1000 0.8 800 0.6

600

0.4

400

0.2

200

Small States Aid Flows

0 1980

Realtive Commodity Price Index

Mitigating the Impacts of Dependent Countries

Composite Price Index 0

1985

1990

1995

2000

Figure 10.3. Composite Relative Commodity Prices and Aid Flows to Small States, 1980–2000. Note : The left vertical axis measures aid ﬂows, while the right vertical axis measures the aggregate relative price index. Aid ﬂows are in terms of 1985 real prices.

1.6

350

1.4

300

1.2

250

1

200

0.8

150

0.6

100

0.4

50

0.2 2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

Real Cotton Price 1989

1988

1987

1986

1985

1984

1983

1982

1981

1980

Real Aid Flows 0

Real Cotton Price Index

400

1990

Real ODA Flows: US$ Million

Source : Real aid ﬂows are authors’ estimate based on data from UNCTAD (2002b), while the relative commodity price index is as constructed in Chapter 3.

0

Figure 10.4. Aid Flows to Mali and Cotton Prices Note : The left vertical axis measures real ODA ﬂows, while the right vertical axis measures the relative commodity price index.

however, have not responded favourably to offset the decline in the relative cotton price and thus compensate for the resultant fall in the purchasing power of exports (see Figure 10.4). Indeed, Figure 10.4 seems to suggest that aid ﬂows to Mali have been positively associated with the real price of cotton.

304

350

1.2

300

1

250

0.8

200 0.6 150 0.4

100

000

0.2

50 PNG Aid Flows

Real Cocoa Price Index

Real Aid flows: US$ Millions

Aid Flows and Commodity Prices

Cocoa Price Index

0

0 1980

1985

1990

1995

2000

Figure 10.5. Aid Flows to Papua New Guinea and the Real Cocoa Price.

180

1.6

160

1.4

140

1.2

120

1.0

100 0.8 80 0.6

60

0.4

40 20

Real ODA Flows

Phosphate Price Index

Real ODA Flows (US$ millions)

Note : The left vertical axis measures real ODA ﬂows, while the relative commodity price index is measured on the right vertical axis.

0.2

Phosphate Prices

0

0.0 1980

1985

1990

1995

2000

Figure 10.6. Aid Flows to Togo and the Real Phosphate Price. Note : The left vertical axis measures the real ODA ﬂows while on the right vertical axis relative commodity price index is measured.

In the second case, that of Papua New Guinea, the cocoa price and aid ﬂows are shown in Figure 10.5. With the sustained decline in cocoa prices since the mid-1980s, Papua New Guinea incurred a cumulative loss of US$562 million during the period 1985–2000 (as estimated in Chapter 6). However, real aid ﬂows to Papua New Guinea declined over the same period—falling from $257.2 million in 1985 to an annual average of US$176.7 million in 1999– 2000. Finally, real ODA ﬂows to Togo, together with the real price of phosphate, the country’s primary commodity export, are plotted in Figure 10.6.6 6 Real prices for other primary commodity exports from Togo—cotton and coffee—have also been persistently weak.

305

Mitigating the Impacts of Dependent Countries Aid ﬂows seem to have no relationship to phosphate price, but in the late 1990s some recovery in the price was accompanied by a substantial decrease in aid inﬂows. While the cases illustrated above might not represent the trends in all individual LDCs, HIPCs, and SVs, there is a general recognition that weakness in relative commodity prices has never been a criterion for aid allocation and there appears to be little evidence that aid ﬂows in recent years have tended to compensate for decreased real commodity prices.7

10.2. A Proposal for the Establishment of an Aid-Financed Diversiﬁcation Fund Declining commodity prices have serious implications for most LDCs, HIPCs and SVs, as it was estimated in Chapter 6 that they caused large foreign exchange losses in terms of reduced purchasing power of primary exports. Increased aid ﬂows to these countries are one way of mitigating this problem. With the end of the EU STABEX scheme, a programme of pure export earnings stabilization has received mixed reviews from donors (Page and Hewitt, 2001).8 Pure compensatory schemes are criticized for two main reasons. First, donors point out that if the goal of a compensatory scheme is temporary stabilization of export revenues, then this may be best attained through the use of the IMF Compensatory Finance Facility (CFF), designed to resolve the temporary balance of payments disequilibria of countries. Second, critics argue that to the extent that grant compensation supports countries for their export earnings, they create adverse incentives and a moral hazard. The moral hazard results as compensatory transfers might discourage developing countries from embarking on the long-term adjustment of their economies given the sustained decline in the real prices of their dominant commodity exports. The present study takes the position that in the face of a secular decline in relative commodity prices, any compensatory scheme for commoditydependent poor countries should differ markedly from the functioning of STABEX (designed as an export earnings stabilization scheme) or from that of the CFF (designed to resolve temporary balance of payment disequilibria). Indeed, the pragmatic policy option in light of the secular declining trend would call for an intensiﬁcation of support measures for export diversiﬁcation and structural change. Our proposal is, therefore, for the establishment of a Joint Diversiﬁcation Fund ( JDF), which would make additional grant transfers available to LDCs, HIPCs, and SVs to enable them to diversify their exports 7 In general, aid ﬂows have focused on growth enhancing and poverty alleviation schemes. But there is also evidence that donors have given aid to ensure that the recipient countries can service their debts (UNCTAD 2002a). 8 Part of the theoretical argument, derived from Newbery and Stiglitz (1981), suggests that stabilization schemes may have very limited welfare beneﬁts.

306

Aid Flows and Commodity Prices away from a few primary commodities. According to the proposal, the JDF would be funded annually by OECD donors working within a multilateral framework. Developing countries satisfying a number of eligibility requirements could subsequently present well-deﬁned diversiﬁcation projects to the scheme for funding consideration. A number of choices and questions emerge in the operationalization of such a compensatory scheme, intended to strengthen the diversiﬁcation process. What is outlined below is meant purely for the purposes of discussion and should not be seen as a deﬁnitive proposal. Clearly, the development of such a programme is beyond the intended scope of this paper and numerous issues would need to be resolved should such a proposal be accepted. Some hypothetical costs to developed countries in supporting the proposed Joint Diversiﬁcation Fund are also calculated, along with the potential burden-sharing arrangement among donors.

10.2.1. Principles in scheme design The proposal below outlines several aspects of the JDF as envisaged by the authors. The implicit assumptions underlying this particular proposal are: . The fund would operate on the principle that contributions and access to the fund should be based upon the net transfer of resources from developing to developed countries that stems from the past change in real commodity prices. The actual resource requirements needed for economic diversiﬁcation in HIPCs, LDCs, and SVs will almost certainly prove to be signiﬁcantly larger than the loss of real purchasing power and hence the fund could only make a partial contribution to the process of diversiﬁcation. . The fund would be aimed at providing resources for coherent private sector projects as well as infrastructural projects for export diversiﬁcation. As diversiﬁcation often requires both public and private sector investments, two windows should ideally be created for such a fund. 10.2.1.1. TIMING OF FUND DISBURSEMENTS The effectiveness of earlier instruments such as STABEX was partly hampered by inefﬁciencies in administration and disbursement of funds from the scheme (Page and Hewitt, 2001; Radetzki, 1990). There were signiﬁcant time lags in the disbursement of funds, so that their release cycle resulted in pro-cyclical, instead of the anticipated counter-cyclical, shocks to recipient economies (Radetzki, 1990). The JDF would not be based upon price volatility but rather upon secular decline in price and would therefore be neutral in terms of price volatility. In our assessment, the disbursement of funds from any compensatory scheme, including the one proposed here, may be designed as ‘retrospective’ or ‘prospective’.

307

Mitigating the Impacts of Dependent Countries A scheme is retrospective in its disbursement cycle if compensation is calculated at the end of each ﬁscal year, and is based on past movements in commodity prices. Under a ‘retrospective’ scheme, compensation is calculated each year based on country-speciﬁc composite commodity price indices, or on a commodity-by-commodity basis (as in the case of STABEX). This is in contrast to ‘prospective’ compensatory schemes where a given volume of funds is committed by donors, and released annually over a deﬁned time span. The proposed JDF combines elements of both retrospective and prospective schemes in that the obligations are known to donor governments a priori but are based upon past price movements. 10.2.1.2. ELIGIBILITY REQUIREMENTS AND GRADUATION CRITERIA The intended beneﬁciaries of any JDF must be clearly speciﬁed if disbursements are to be effectively targeted. Previous compensatory schemes targeted various groups based on the objectives of the scheme—whether it aimed at redressing balance of payments difﬁculties (e.g. CFF) or ensuring export earnings stabilization (e.g. STABEX). In cases where commodity-speciﬁc stabilization is the goal, transfers to each country are made on a sectoral basis and are not tied to overall primary commodity receipts or to the export performance of other sectors of the economy (such as manufacturing or services). Such a scheme presents mixed blessings as the eligibility requirements may predict disbursements to a country experiencing an unusual adverse shock in one commodity, when its net primary commodity export earnings are fairly stable (or even increasing). Disbursements from such a fund may, therefore, be heavily skewed, with disproportionate shares being allocated to a few countries.9 As presently envisaged, the eligibility criteria for the JDF must be such as to result in the inclusion of a set of poor countries relying on primary commodities for a large fraction of their total export earnings. In the design of the JDF, we restrict our analysis to LDCs, HIPCs and small vulnerable states. Countries would have to decide whether all commodity-exporting countries in the group would be eligible to take part in the programme, or only those that had achieved a threshold dependence ratio on commodities. Should countries choose any above zero threshold of commodity dependence for the JDF, as presented in Table 10.1 below, it would also be necessary to devise an appropriate graduation arrangement for those entering or exiting the included group, e.g. LDCs, HIPCs, and countries with a 25 per cent dependence threshold.

9 ˆ te d’Ivoire, Ghana and This was the curse of STABEX where four countries (Senegal, Co Sudan) alone received over 40 per cent of fund disbursements for the period 1975–84 (Hewitt, 1987).

308

Aid Flows and Commodity Prices 10.2.1.3. CALCULATING THE ANNUAL CONTRIBUTIONS According to our proposal, the JDF is to be initially capitalized by donor contributions and will require annual replenishments equivalent in total to the net transfers to developed countries from the real price decrease of exports over the reference period, which in the case illustrated below is six years. Ultimately, the actual size of an individual country’s contributions will remain a political decision subject to negotiation, though an attempt to develop a country-speciﬁc approach is considered below. There exist two main choices for calculating the contribution to the JDF scheme. First, one might calculate annual commodity earning losses of eligible countries based on trend or average decline in commodity prices.10 However, as discussed in Chapter 6, computed trend or average rates are sensitive to the choice of the base year in the sample, and to the reference period examined. Calculations of payments based on trend estimates are also tedious in practice, as they require detailed knowledge of a country’s export volumes.11 An alternative method is to base contributions to the fund on real export earning losses by countries with reference to a set of past real prices. This is the same approach as that adopted in Chapter 6 in quantifying terms of trade losses of commodity-dependent countries. A practical question arises as to which set of real prices should be chosen as the reference year for calculation. A set of (relatively) recent prices may be required, but a scheme sensitive only to annual movements in real prices may be undesirable.12 The desired payments into the JDF must therefore be calculated relative to a past base year. The choice of a base year in the mid-1980s may be desirable as it precedes the major commodity price collapse of the late 1980s. Other suggestions may be to construct price indices based on a moving average index of real commodity prices, calculated over a number of years. 10.2.1.4. POOLED OR COUNTRY-SPECIFIC ENVELOPES Rather than individual country allocations, we recommend that the scheme operates as a common pool arrangement (Kanbur et al., 1999), and is annually replenished by donors based on past known commodity price trends and resulting 10 Besides, it might be argued that for practical implementation, the trend rate for the most recent sample (for example, since the 1980s) is more relevant. Estimation of such a recent trend decline rate is likely to be very low, as real prices for most commodities had already fallen to a very low level by the 1980s. 11 It was also pointed out in Chapter 6 that calculation of losses based on the trend rate of decline would require ﬁnding out a representative volume of exports, which might not be a straightforward task. 12 A compensation scheme based solely on annual movements in prices is undesirable for two reasons. First, a temporary upward bulge in commodity prices may not disburse any returns from the fund although current prices may be severely depressed compared with prices over a broader time period (for example, since the 1960s). Second, when the current prices are already so depressed, a scheme triggered solely by year-to-year price changes is likely severely to underestimate the export earning shortfalls.

309

Mitigating the Impacts of Dependent Countries loss of purchasing power of exports incurred by LDCs, HIPCs and SVs. Beneﬁciaries’ maximum envelopes could be based upon share of total real losses in the commodities covered. However, this would imply a country-speciﬁc envelope and a homogeneous capacity to develop viable and coherent projects. Where individual countries are unable to generate projects, the pooled fund nature would allow individual LDCs, HPICs, and SVs to draw on the fund in amounts over and above individual country losses in the reference period. 10.2.1.5. AID FUNGIBILITY AND MORAL HAZARD A contentious debate exists on the fungibility of aid (Cassen et al., 1994) and its consequences for aid effectiveness (Devarajan and Swaroop, 1998). Critics argue that to the extent that aid is fungible, speciﬁc development goals may not be attained as additional resources only permit recipient governments to reallocate budget expenditure in favour of immediate consumption needs or other lower priority projects. Untied aid is fully fungible, as it is readily absorbed by the recipient government’s most pressing social expenditure needs, or used to accomplish balance of payments needs such as debt servicing. Even in cases when aid is tied, critics argue that aid is fungible as additional donor ﬁnance simply ‘crowds out’ similar investments which would have been made by recipient governments. The basic argument is that aid relaxes a government’s budget constraint, enabling it to reallocate resources. While these arguments may be valid for large-donor projects targeting social expenditure, they may not be prominent in the case of the present proposal for a diversiﬁcation fund, if only because the country-speciﬁc resources in the JDF are unlikely to be of such an order of magnitude as to induce crowding out. The problem of the fungibility of untied aid can be addressed by setting stringent requirements on resource use. We propose that the JDF grants should be tied solely to diversiﬁcation projects, and disbursed to recipient governments presenting coherent strategies for funding with a view to reducing dependence on a limited range of commodities. To the extent that fungibility issues arise, they are more correctly of concern with regard to donor contribution rather than to recipient use of resources in such a programme. Indeed, if the JDF were accepted, it would be essential for funding to come from additional resources as real aid budgets are generally not expanding and hence the programme could be developed at the expense of other potentially important programmes, such as those combating HIV/AIDS. The JDF may present adverse incentives for diversiﬁcation. In principle, this may also mean that countries which have reduced their dependence on primary exports via successful diversiﬁcation schemes might be discouraged from continuing their efforts. There was certainly some evidence of this under STABEX when it operated as a scheme involving the transfer of public funds

310

Aid Flows and Commodity Prices to the coffers of recipient beneﬁciaries. However, we ﬁnd moral hazard concerns to be insigniﬁcant in the present case. As argued in the case of aid fungibility, the relatively small size of annual transfers, and the tying of funds to speciﬁc diversiﬁcation projects, greatly minimize the scope for moral hazard behaviour. In the proposed design, triggering compensation into the joint fund does not guarantee automatic fund disbursements, as countries need to present planned diversiﬁcation projects.

10.2.2. Hypothetical costs to developed countries of the diversiﬁcation fund 10.2.2.1. COSTS OF SCHEME Let us consider an aid-ﬁnanced scheme to compensate for foreign exchange losses incurred by LDCs, HIPCs, and SVs due to declining commodity prices but tied to diversiﬁcation projects. The diversiﬁcation fund can be based on either a full or partial adjustment for primary commodity export earning losses. Another factor is the determination of the threshold level of dependence on commodities for export earnings that will make countries eligible for the diversiﬁcation support. We simulate the potential costs of a number of schemes on the basis of 50 per cent compensation and various threshold dependence levels, assuming that they had been in operation for the period 1995–2000. Table 10.1 summarizes the results of hypothetical cost estimates. The size of the proposed fund varies signiﬁcantly depending on the commodity dependence threshold and rate of contribution for commodity-related foreign exchange losses. The second column shows that a scheme which is aimed at making up for 50 per cent real foreign exchange loss by individual LDCs, HIPCs, and SVs, but without needing any dependence threshold, would have required US$1.9 billion (in 1984–86 prices) in 1995.13 However, if the size of the fund is to depend on foreign exchange losses of those countries that receive at least 50 per cent of their export earnings from commodities, the comparable ﬁgure would be about US$1.3 billion (in 1984–86 prices) in 1995, which increases to approximately US$2 billion in 2000 (column 4 in Table 10.1). Estimation of the fund size based on 25 per cent and 75 per cent dependence on commodities is also presented in Table 10.1. Corresponding to aggregate results as summarized in Table 10.1, details of the size of grants available to each country are presented in Appendix 10.6, paras. 10.1–10.4.

13 Note that this ﬁgure is higher than 50 per cent of the foreign exchange loss estimated for LDCs, HIPCs and SVs in Chapter 6. This is because the net loss calculated in Chapter 6 was obtained by summing over both negative (losses) and positive (terms of trade gains) ﬁgures. But, for the exercises in Table 10.1, the compensation is based on countries that have incurred only foreign exchange losses.

311

Mitigating the Impacts of Dependent Countries Table 10.1. Cost Estimates for a Joint Diversiﬁcation Fund for LDCs, HIPCs, and Small States (US$ million in 1984–86 prices) Size of Fund with 50% Adjustment for Foreign Exchange Loss from Commodities

Year

No dependence threshold

25% dependence threshold

50% dependence threshold

75% dependence threshold

1995 1996 1997 1998 1999 2000

1932.06 2666.40 3212.79 3899.73 4413.30 4385.20

1790.68 2562.96 3007.87 3458.47 3990.70 3965.00

1292.22 1815.98 1944.22 2152.61 2391.86 2053.54

530.23 722.31 704.52 636.34 694.99 712.46

20509.47

18775.67

11650.42

4000.84

TOTAL

Note: The dependence threshold in columns 2, 3 and 4 implies that countries are only considered for the fund if their earnings from commodities account for, respectively, 25, 50, and 75 per cent of their merchandise exports. Estimates in column 1 (i.e. with no dependence threshold) does not consider individual countries’ dependence on commodities for exports. Source: Authors’ estimates.

10.2.2.2. BURDEN SHARING AMONG DONORS The establishment of the JDF will involve costs for donor countries. For the sixyear period examined, 1995–2000, estimated costs for the scheme total approximately US$20 billion in 1984–86 prices, i.e. about US$25 billion in 2000 nominal prices (under the no dependence threshold condition as presented in column 2 of Table 10.2).14 This ﬁgure drops to US$11.6 billion in real terms or US$14 billion in nominal 2000 dollars if a 50% threshold is employed. We now consider hypothetical burden-sharing arrangements among OECD’s 22 Development Assistance Committee (DAC) members. In some recent studies, donor contributions are calculated according to their ability to pay (Addison et al., 2003), which is computed as the relative share of an individual country’s GDP (or GNI) of the combined DAC total. However, as many DAC members do not meet the UN stipulated ODA target, a scheme requiring more contributions from donors with relatively low ODA/GNP ratios may be desirable. Consequently, we illustrate a funding arrangement where half the required funds are contributed by donors based on the relative sizes of their GNI, and the other half depending on ODA/GNP ratios. For the second half of the contribution, following Berlage et al. (2003), individual country contributions, Ci , to the fund can be computed as: Ci ¼ (0:7 Łi )Gi

14 The conversion from 1984–86 real prices to 2000 current US dollars is based on the changes in the manufacturing export unit value index of developed market economy countries over the same period.

312

Aid Flows and Commodity Prices where Gi is the GNI of donor i, with f chosen such that Ci contributions sum up to $25 billion (in nominal US dollars).15 Results for this hypothetical burden-sharing arrangement are presented in Tables 10.2 and 10.3. From Table 10.2 it is observed that based on the share of combined DAC GNI, the US has to make the largest contribution, followed by the EU. Since the US ODI/ GNI ratio of 0.1 is much lower than the UN stipulated targets, it will also have to make a relatively high contribution to the JDF. Denmark, Luxembourg, the Netherlands, Norway and Sweden have negative ﬁgures for share based on ODA targets. This is because the ODA that these countries currently provide amounts to more than the UN target of 0.7 per cent of their GNI. On the whole, about 75 per cent of the assistance should come from the US and the EU.16 Table 10.3 estimates the hypothetical annual contribution of the donor countries, had the programme been operated in 1995–2000. The US, the EU and Japan would have been required to provide, respectively, US$1.9 billion, US$1.2 billion and US$0.8 billion. These costs fall to US$1 billion, US$0.65 billion and US$0.5 billion with a 50% dependence threshold. Even after the contribution to the diversiﬁcation fund, for most countries the ODI/GNI share would be much lower than the UN target. In the case of the US and the EU, the ODI/GNI ratio would increase only marginally—by just 0.2 percentage points. On the whole, funding of the scheme would require an annual increase of donor ODA ﬂows of 9.31 per cent. It might be of interest to compare the estimated EU funding under the current scheme with that under the STABEX programme. Total contributions from EU members in this hypothetical arrangement are estimated at US$7.2 billion for the six-year period, implying an annual average contribution of about US$1.2 billion (or about !1.07 billion). Under the Lome´ IV negotiations, STABEX funds amounted to !1.8 billion for the ﬁnal ﬁve-year period of 1995–2000—i.e. annual payments of !360million (Page and Hewitt, 2001). Therefore, for EU member countries, the current scheme of 50 per cent compensation for LDCs, HIPCs and SVs with no commodity dependence threshold requires three times the support provided under STABEX. With a 50% dependence threshold, the EU contribution would be halved. Another way of calculating the donors’ contribution to the JDF might be on the basis of some combination of their GNI and their imports of primary commodities from LDCs, HIPCs and SVs. Data on imports of primary commodities from the recipient countries by DAC members are not available, so it was not possible to undertake such an exercise.17 However, given the general 15 In this present exercise f was calculated as 0.0013 for the set of estimates presented in Tables 10.2 and 10.3. 16 The share of the US in the JDF fund is calculated to be 46 per cent, while the corresponding ﬁgure for the EU is 29 per cent. The third largest contribution is estimated for Japan, which is about 19 per cent. 17 Data on DAC member countries’ total imports of primary commodities are, however, available. But a large portion of these imports are supplied by other countries not included in the sample of LDCs, HIPCs, and SVs.

313

Mitigating the Impacts of Dependent Countries Table 10.2. Hypothetical Burden Sharing among Donors Contribution to Joint Diversiﬁcation Fund (US$ million) Share of Share based GNI in 2000 Combined ODA in 2000 Share based on ODA (US$ million) DAC GNI (US$ million) ODA/GNI on GNI targets Total Australia Austria Belgium Canada Denmark Finland France Germany Greece Ireland Italy Japan Luxembourg Netherlands New Zealand Norway Portugal Spain Sweden Switzerland United Kingdom United States TOTAL DAC Of which : EU Members

370209.09 186121.01 229121.48 684489.93 156876.5 119307.19 1284764.47 1862363 111916 79335.36 1074281.77 4807580.71 17314.07 374644.53 45017.01 158010 103795.03 552060 224515.64 258364.53 1417780.81 9928500

0.015 0.008 0.010 0.028 0.007 0.005 0.053 0.077 0.005 0.003 0.045 0.200 0.001 0.016 0.002 0.007 0.004 0.023 0.009 0.011 0.059 0.413

987.14 423.27 819.66 1743.6 1664.18 370.84 4104.71 5030 226 234.82 1376.26 13507.96 122.97 3134.78 113.22 1263.56 270.62 1194.82 1798.95 890.37 4501.26 9954.89

0.267 0.227 0.358 0.255 1.061 0.311 0.319 0.270 0.202 0.296 0.128 0.281 0.710 0.837 0.252 0.800 0.261 0.216 0.801 0.345 0.317

192.45 96.75 119.10 355.82 81.55 62.02 667.86 968.11 58.18 41.24 558.44 2499.12 9.00 194.75 23.40 82.14 53.96 286.98 116.71 134.31 737.00

175.01 95.95 85.54 332.47 61.75 50.65 533.27 873.38 60.80 34.96 670.18 2197.51 0.19 55.88 22.02 17.18 49.74 291.21 24.80 100.16 591.58

367.45 192.70 204.65 688.29 19.80 112.67 1201.13 1841.50 118.98 76.21 1228.62 4696.63 8.81 138.87 45.43 64.96 103.69 578.19 91.91 234.46 1328.59

0.100

5161.12

6495.36

11656.48

24046368.13

1.000

53 734

0.223

12500.00

12500.00

25000.00

7794196.86

0.324

25 273

0.324

4051.65

3194.66

7246.31

Note: The compensation scheme as presented in this table is based on 50 per cent compensation for the foreign exchange loss incurred by all LDCs, HIPCs, and SVs. The estimates of foreign exchange loss in 1984–86 prices come from Chapter 4, but have been converted into nominal dollars. The present hypothetical exercise considers a funding arrangement where half the required funds are contributed by donors on the basis of the relative size of their GNI, and the other half depend on ODA/GNP ratios. Source: Authors’ estimates based on data on GNI and ODA ﬂows from World Bank (2002).

illustrations in Tables 10.2 and 10.3, the burden-sharing arrangement among donors under alternative scenarios can be readily undertaken when the information becomes available.

10.3. Conclusions The secular decline in relative commodity prices requires commoditydependent countries to pursue growth strategies that would reduce their reliance on a narrow set of primary commodity exports. Therefore, export diversiﬁcation provides a useful forward-looking growth strategy. Our proposal in this chapter is for the establishment of a multilateral fund that provides a means of supporting diversiﬁcation projects in commodity-dependent LDCs, HIPCs, and SVs.

314

Aid Flows and Commodity Prices Table 10.3. ODA/GNI Positions of Donors after the Hypothetical Contributions to the Joint Diversiﬁcation Fund Annual Percentage Increase in Annual ODA (based ODA (2000) Contribution on 2000 values) Australia Austria Belgium Canada Denmark Finland France Germany Greece Ireland Italy Japan Luxembourg Netherlands New Zealand Norway Portugal Spain Sweden Switzerland United Kingdom United States TOTAL DAC Of which EU Members

987.14 423.27 819.66 1743.6 1664.18 370.84 4104.71 5030 226 234.82 1376.26 13507.96 122.97 3134.78 113.22 1263.56 270.62 1194.82 1798.95 890.37 4501.26 9954.89 53733.88 25 273

61.24 32.12 34.11 114.71 3.30 18.78 200.19 306.92 19.83 12.70 204.77 782.77 1.47 23.15 7.57 10.83 17.28 96.36 15.32 39.08 221.43 1942.75 4166.67 1207.72

6.20 7.59 4.16 6.58 0.20 5.06 4.88 6.10 8.77 5.41 14.88 5.79 1.19 0.74 6.69 0.86 6.39 8.07 0.85 4.39 4.92 19.52 7.75 4.78

Current Additional Ratio of increase Total ODA/GNI in ODA/GNI ODA/GNI (expressed (expressed (expressed in percent- in percentage in percentage age points) points) points) 0.267 0.227 0.358 0.255 1.061 0.311 0.319 0.270 0.202 0.296 0.128 0.281 0.710 0.837 0.252 0.800 0.261 0.216 0.801 0.345 0.317 0.100 0.223 0.324

0.017 0.017 0.015 0.017 0.002 0.016 0.016 0.016 0.018 0.016 0.019 0.016 0.008 0.006 0.017 0.007 0.017 0.017 0.007 0.015 0.016 0.020 0.017 0.015

0.283 0.245 0.373 0.271 1.063 0.327 0.335 0.287 0.220 0.312 0.147 0.297 0.719 0.843 0.268 0.807 0.277 0.234 0.808 0.360 0.333 0.120 0.241 0.340

Source: Authors’ estimates.

Before concluding the chapter, it is worth pointing out two important issues. First, the fund for diversiﬁcation should not be considered as a substitute for other aid ﬂows to the recipient countries. In fact, it must be supplemented by regular and increased aid ﬂows, which are critical for supporting overall growth and development in poor countries. Last, but not least, setting up a diversiﬁcation fund does not guarantee success in attracting local or foreign investment in diversiﬁcation projects; a host of internal economic reforms are also needed in many commodity-dependent countries to establish an appropriate investment climate. An export diversiﬁcation fund cannot be a panacea for the problems of commodity-dependent developing countries. However, if it is based on additional aid resources and is combined with sound economic policies, it would make a valuable contribution to redress the negative development impact of the secular decline in real commodity prices.

315

Appendix 10.1. Country Allocations from JDF with No Dependence Threshold (with Compensation 50 per cent of Foreign Exchange Loss from Commodities) (Dependence Threshold 0.0)

COUNTRY Afghanistan Antigua and Barbuda Angola Bangladesh Barbados Belize Benin Bhutan Bolivia Botswana Burkina Faso Burundi Cambodia Cameroon Cape Verde CAR Chad Comoros Congo ˆ te D’Ivoire Co DRC

Dep. Index

1995

1996

1997

1998

1999

2000

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Dep. Index

Dep. Index

Dep. Index

Dep. Index

Dep. Index

0.56 0.04

8.93

0.82 0.07

11.07

0.60 0.06

9.37

0.48 0.06

6.53

0.70 0.05

7.42 0.09

0.53 0.07

0.01 0.14 0.30 0.92 0.48 0.18 0.64 0.14 0.52 0.93 0.26 0.67 0.27 0.49 0.61 0.59 0.12 0.71 0.91

4.81

0.01 0.13 0.37 0.90 0.36 0.19 0.68 0.12 0.50 0.93 0.25 0.64 0.25 0.42 0.55 0.27 0.07 0.67 0.78

4.15

0.01 0.12 0.35 0.84 0.51 0.15 0.74 0.10 0.62 1.00 0.17 0.54 0.24 0.44 0.64 0.59 0.10 0.62 0.76

5.45

0.01 0.12 0.32 0.89 0.47 0.16 0.68 0.15 0.67 1.00 0.10 0.57 0.23 0.66 0.56 0.74 0.10 0.64 0.88

10.32

0.01 0.11 0.30 0.96 0.32 0.15 0.67 0.11 0.53 1.00 0.08 0.59 0.22 0.58 0.46 0.57 0.08 0.58 0.68

0.84

0.01 0.11 0.29 0.88 0.35 0.12 0.64 0.12 0.62 0.78 0.06 0.52 0.42 0.60 0.51 0.41 0.05 0.57 0.64

2.75 216.35 5.23 28.78 32.13 0.19 12.88 8.04 1.27 14.02 386.23 4.63

4.56 187.16 4.73 25.69 37.85 18.51 0.16 15.63 6.49 0.93 3.74 495.49 20.79

5.53 201.50 17.80 65.20 40.09 0.25 27.28 8.06 1.18 17.48 510.25 28.77

6.47 205.01 34.04 57.03 34.17 16.16 0.18 57.44 12.54 0.84 41.24 522.14 64.43

6.98 7.11 201.60 47.52 78.92 34.23 113.85 0.20 66.16 9.85 0.87 15.92 474.49 29.72

8.18 0.14

17.84 8.25 2.70 6.70 211.73 42.92 80.14 31.57 0.42 89.81 13.00 0.74 419.23 24.83

Djibouti Dominica Equatorial Guinea Eritrea Ethiopia Fiji Gabon Gambia Ghana Grenada Guinea Guinea-Bissau Guyana Haiti Honduras Jamaica Kenya Kiribati Lao PDR Lesotho Liberia Madagascar Malawi Maldives Mali Mauritania Mauritius Mozambique Myanmar Nepal Nicaragua Niger Papua New Guinea Rwanda Samoa Sao Tome and Principe Senegal

0.28 0.55 0.31

0.26

0.32 0.97 0.51 0.12 0.92 0.49 0.80 0.74 0.92 0.66 0.34 0.50 0.71 0.66 0.95 0.47 0.08 0.06 0.86 0.89 0.74 0.61 0.59 0.30 0.81 0.91 0.17 0.82 0.72 0.61

0.22 23.18 71.38 63.12

9.10 0.46 11.81 29.05

21.07 0.95 25.51 57.71 2.95 6.75

20.88 112.79 8.71 22.58 45.54

0.20 0.48 0.26

0.35

0.18 0.95 0.44 0.10 0.91 0.69 0.81 0.64 0.91 0.71 0.34 0.45 0.74 0.61 0.96 0.38 0.05 0.08 0.95 0.80 0.83 0.73 0.61 0.30 0.62 0.91 0.17 0.92 0.58 0.58

0.37 31.22 7.05 43.45

0.74 0.86 0.73

0.30

0.84 0.91 0.88

0.44

8.23

0.46

0.17 0.53 0.13

0.31

0.22 35.02 11.45 56.83

40.98 158.55 15.55 50.16 55.30 31.54

0.06 0.90 0.40 0.12 1.00 0.58 0.79 0.61 0.57 0.55 0.28 0.44 0.77 0.62 0.95 0.19 0.05 0.11 0.88 0.94 0.77 0.53 0.65 0.29 0.74 0.73 0.34 0.83 0.66 0.56

0.20 0.26 5.58

95.32

10.12 11.40 54.42

0.13 28.64 1.91 14.77 51.99 2.45 7.43

0.30 0.38 0.14

0.76

0.32 59.57 8.36 188.42

51.11 203.86 44.92 63.36 59.63 55.47

0.12 0.95 0.41 0.20 0.57 0.47 0.79 0.52 0.71 0.72 0.20 0.53 0.76 0.73 0.91 0.15 0.04 0.12 0.88 0.72 0.76 0.56 0.52 0.28 0.60 0.61 0.15 0.62 0.74 0.59

0.65 0.93 0.66

0.64 0.36

0.84 0.94 0.50

0.47

3.57

0.46

76.98

8.83 11.64

0.27 21.86 2.26 15.13 51.67 4.35 13.46

0.69 0.43 0.14

3.87 1.01

0.62 0.44 0.08

0.80 78.16 6.90 199.92

43.20 241.97 25.93 42.62 79.38 42.24

0.32 0.90 0.37 0.20 0.99 0.54 0.57 0.53 0.77 0.63 0.13 0.42 0.81 0.64 0.81 0.20 0.04 0.16 0.92 0.77 0.61 0.43 0.52 0.26 0.51 0.58 0.13 0.74 0.85 0.52

44.71 267.74 32.86 65.98 84.26 33.15

0.39 0.84 0.42 0.15 0.96 0.68 0.49 0.51 0.69 0.70 0.16 0.55 0.81 0.64 0.68 0.19 0.03 0.29 0.53 0.82 0.54 0.47 0.60 0.25 0.48 0.41 0.08 0.72 0.83 0.51

0.71 0.39

0.80 0.92 0.74

0.62 0.72

0.80 0.91 0.95

0.32 1.12

0.45

12.36

0.48

33.32

20.62

12.95 0.12 10.96

0.26 21.00 2.58 16.40 85.75 4.35 7.79

73.69 1.11 23.51 18.16 8.94 2.06 43.76 0.27 25.20 0.56 5.93 16.45 58.47 3.16 26.04

1.12

5.57 214.49 112.00 1.19 35.87 20.00 9.76 25.44 50.58 0.21 24.94 0.69 16.20 12.57 67.08 3.72 25.26 4.99 37.11 260.26 24.78 78.96 98.07 26.59

(Continued )

Appendix 10.1. (Continued )

COUNTRY Seychelles Sierra Leone Solomon Islands Somalia St Kitts and Nevis St Lucia St Vincent Sudan Suriname Swaziland Togo Tonga Trinidad and Tobago Tuvalu Uganda United Rep. of Tanzania Vanuatu Vietnam Yemen Zambia TOTAL

Dep. Index

1995

1996

1997

1998

1999

2000

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

0.46 0.77 0.98

1.97

0.59 0.86

5.30

0.52 0.99 0.91 0.96 0.34 0.52 0.81 0.08 0.48 0.89 0.69 0.73 0.38 0.05 0.80

Dep. Index 0.30 0.52 0.97 0.56 0.61

19.41

0.72 0.80 0.92 0.95 0.42 0.50 0.81 0.08

125.08 100.73

27.84 26.60 13.18 24.70

304.37 14.12 1932.06

2.09 4.14 6.29

Dep. Index 0.63 0.58 0.99 0.70 0.76

42.44 22.74 23.46 34.13 0.21 6.73

0.61 0.76 0.96 0.79 0.44 0.51 0.92 0.10

0.32 0.84 0.69

0.01 218.81 115.02

0.79 0.33 0.03 0.77

0.51 559.93 17.73 56.03 2666.40

5.49 9.85 5.52

Dep. Index 0.74 0.34 1.00 0.48 0.57

6.93 6.40 3.61

Dep. Index

10.48

3.44

0.65 0.31 0.91

0.62 0.45

6.50 0.51

0.63 0.39

9.86 0.56

1.66 1.64 66.58 30.56 31.77 14.57 0.36 9.51

0.84 0.81 0.21 0.91 0.49 0.48 0.87 0.06

1.84 1.84

0.76 0.32 0.94

8.94

Dep. Index

2.51

37.93 20.17 27.83 28.60 0.46 10.00

0.62 0.85 0.86 0.92 0.49 0.62 0.88 0.11

56.82 22.25 36.62 18.68 0.27 33.57

0.68 0.78 0.46 0.88 0.45 0.49 0.88 0.09

0.71 0.83 0.78

0.02 219.51 78.60

0.67 0.85 0.89

0.01 209.05 137.44

0.25 0.90 0.69

0.01 257.85 102.38

0.49 0.91 0.58

0.01 367.27 114.94

0.86 0.32 0.03 0.96

1.40 936.96 26.83 72.17

1.00 0.35 0.06 0.56

1.36 1235.88 32.05 113.58

0.95 0.30 0.03 0.57

0.81 1426.40 33.76 179.95

0.91 0.26 0.03 0.69

0.57 1470.40 46.36 123.65

3212.79

3899.73

4413.30

32.59 31.77 19.84 0.28

4385.19

Appendix 10.2. Country Allocations from JDF with No Dependence Threshold (with Compensation 50 per cent of Foreign Exchange Loss from Commodities) (Dependence Threshold 0.25)

COUNTRY

Dep. Index

Afghanistan Antigua and Barbuda Angola Bangladesh Barbados Belize Benin Bhutan Bolivia Botswana Burkina Faso Burundi Cambodia Cameroon Cape Verde CAR Chad Comoros Congo ˆ te D’Ivoire Co DRC Djibouti

0.56 0.04 0.01 0.14 0.30 0.92 0.48 0.18 0.64 0.14 0.52 0.93 0.26 0.67 0.27 0.49 0.61 0.59 0.12 0.71 0.91 0.28

1995

1996

1997

1998

1999

2000

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

8.93

216.35

28.78 32.13 0.19 12.88 8.04 1.27 386.23 4.63 0.26

Dep. Index 0.82 0.07 0.01 0.13 0.37 0.90 0.36 0.19 0.68 0.12 0.50 0.93 0.25 0.64 0.25 0.42 0.55 0.27 0.07 0.67 0.78 0.20

11.07

187.16

25.69 37.85 18.51 15.63 6.49 0.93 495.49 20.79

Dep. Index 0.60 0.06 0.01 0.12 0.35 0.84 0.51 0.15 0.74 0.10 0.62 1.00 0.17 0.54 0.24 0.44 0.64 0.59 0.10 0.62 0.76 0.17

9.37

201.50

65.20

27.28 8.06 1.18 510.25 28.77

Dep. Index 0.48 0.06 0.01 0.12 0.32 0.89 0.47 0.16 0.68 0.15 0.67 1.00 0.10 0.57 0.23 0.66 0.56 0.74 0.10 0.64 0.88 0.30

6.53

205.01

57.03 16.16 57.44 12.54 0.84 522.14 64.43 0.76

Dep. Index 0.70 0.05 0.01 0.11 0.30 0.96 0.32 0.15 0.67 0.11 0.53 1.00 0.08 0.59 0.22 0.58 0.46 0.57 0.08 0.58 0.68 0.69

7.42

6.98

201.60

78.92 113.85 66.16 9.85 0.87 474.49 29.72 3.87

Dep. Index 0.53 0.07 0.01 0.11 0.29 0.88 0.35 0.12 0.64 0.12 0.62 0.78 0.06 0.52 0.42 0.60 0.51 0.41 0.05 0.57 0.64 0.62

8.18

8.25 2.70 211.73

80.14

0.42 89.81 13.00 0.74 419.23 24.83

(Continued )

Appendix 10.2. (Continued )

COUNTRY

Dep. Index

Dominica Equatorial Guinea Eritrea Ethiopia Fiji Gabon Gambia Ghana Grenada Guinea Guinea-Bissau Guyana Haiti Honduras Jamaica Kenya Kiribati Lao PDR Lesotho Liberia Madagascar Malawi Maldives Mali Mauritania Mauritius Mozambique Myanmar

0.55 0.31 0.32 0.97 0.51 0.12 0.92 0.49 0.80 0.74 0.92 0.66 0.34 0.50 0.71 0.66 0.95 0.47 0.08 0.06 0.86 0.89 0.74 0.61 0.59 0.30 0.81 0.91

1995

1996

1997

1998

1999

2000

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

0.22 23.18

63.12

9.10 0.46 11.81 29.05

21.07

25.51 57.71 2.95 6.75

20.88 112.79

Dep. Index 0.48 0.26 0.18 0.95 0.44 0.10 0.91 0.69 0.81 0.64 0.91 0.71 0.34 0.45 0.74 0.61 0.96 0.38 0.05 0.08 0.95 0.80 0.83 0.73 0.61 0.30 0.62 0.91

31.22 7.05

95.32

10.12 11.40 54.42

0.13 28.64

14.77 51.99 2.45 7.43

40.98 158.55

Dep. Index 0.53 0.13 0.06 0.90 0.40 0.12 1.00 0.58 0.79 0.61 0.57 0.55 0.28 0.44 0.77 0.62 0.95 0.19 0.05 0.11 0.88 0.94 0.77 0.53 0.65 0.29 0.74 0.73

35.02 11.45

76.98

8.83 11.64

0.27

15.13 51.67 4.35 13.46

51.11 203.86

Dep. Index 0.38 0.14 0.12 0.95 0.41 0.20 0.57 0.47 0.79 0.52 0.71 0.72 0.20 0.53 0.76 0.73 0.91 0.15 0.04 0.12 0.88 0.72 0.76 0.56 0.52 0.28 0.60 0.61

59.57 8.36

20.62

12.95 0.12

0.26

16.40 85.75 4.35 7.79

43.20 241.97

Dep. Index 0.43 0.14 0.32 0.90 0.37 0.20 0.99 0.54 0.57 0.53 0.77 0.63 0.13 0.42 0.81 0.64 0.81 0.20 0.04 0.16 0.92 0.77 0.61 0.43 0.52 0.26 0.51 0.58

1.01 0.80 78.16 6.90

73.69 1.11 23.51 18.16 2.06 43.76 0.27

16.45 58.47 3.16 26.04

44.71 267.74

Dep. Index 0.44 0.08 0.39 0.84 0.42 0.15 0.96 0.68 0.49 0.51 0.69 0.70 0.16 0.55 0.81 0.64 0.68 0.19 0.03 0.29 0.53 0.82 0.54 0.47 0.60 0.25 0.48 0.41

1.12

5.57

112.00 1.19 35.87 20.00 25.44 50.58 0.21

16.20 12.57 67.08 3.72 25.26 4.99 37.11 260.26

Nepal Nicaragua Niger Papua New Guinea Rwanda Samoa Sao Tome and Principe Senegal Seychelles Sierra Leone Solomon Islands Somalia St Kitts and Nevis St Lucia St Vincent Sudan Suriname Swaziland Togo Tonga Trinidad and Tobago Tuvalu Uganda United Rep. of Tanzania Vanuatu Vietnam Yemen Zambia TOTAL

0.17 0.82 0.72 0.61 0.74 0.86 0.73 0.44 0.46 0.77 0.98 0.59 0.86 0.52 0.99 0.91 0.96 0.34 0.52 0.81 0.08 0.48 0.89 0.69 0.73 0.38 0.05 0.80

22.58 45.54

0.30 8.23 1.97

5.30

27.84 26.60 13.18 24.70

125.08 100.73 304.37

1790.68

0.17 0.92 0.58 0.58 0.84 0.91 0.88 0.46 0.30 0.52 0.97 0.56 0.61 0.72 0.80 0.92 0.95 0.42 0.50 0.81 0.08 0.32 0.84 0.69 0.79 0.33 0.03 0.77

50.16 55.30 31.54 0.20 0.26 5.58 2.09 4.14 6.29

42.44 22.74 23.46 34.13 0.21 0.01 218.81 115.02 0.51 559.93 56.03 2562.96

0.34 0.83 0.66 0.56 0.65 0.93 0.66 0.47 0.63 0.58 0.99 0.70 0.76 0.61 0.76 0.96 0.79 0.44 0.51 0.92 0.10 0.71 0.83 0.78 0.86 0.32 0.03 0.96

44.92 63.36 59.63 55.47 0.64 0.36 3.57 5.49 9.85 5.52

37.93 20.17 27.83 28.60 0.46 0.02 219.51 78.60 1.40 936.96 72.17 3007.87

0.15 0.62 0.74 0.59 0.84 0.94 0.50 0.46 0.74 0.34 1.00 0.48 0.57 0.62 0.85 0.86 0.92 0.49 0.62 0.88 0.11 0.67 0.85 0.89 1.00 0.35 0.06 0.56

42.62 79.38 42.24 0.71 0.39 6.93 6.40 3.61

56.82 22.25 36.62 18.68 0.27 0.01 209.05 137.44 1.36 1235.88 113.58 3458.47

0.13 0.74 0.85 0.52 0.80 0.92 0.74 0.45 0.76 0.32 0.94 0.62 0.45 0.68 0.78 0.46 0.88 0.45 0.49 0.88 0.09 0.25 0.90 0.69 0.95 0.30 0.03 0.57

65.98 84.26 33.15 0.62 0.72 12.36 8.94 3.44 6.50 0.51 1.66 1.64 66.58 30.56 31.77 14.57 0.36

257.85 102.38 0.81 1426.40 179.95 3990.69

0.08 0.72 0.83 0.51 0.80 0.91 0.95 0.48 0.65 0.31 0.91 0.63 0.39 0.84 0.81 0.21 0.91 0.49 0.48 0.87 0.06 0.49 0.91 0.58 0.91 0.26 0.03 0.69

78.96 98.07 26.59 0.32 1.12 33.32 10.48 2.51 9.86 0.56 1.84 1.84 32.59 31.77 19.84 0.28 0.01 367.27 114.94 0.57 1470.40 123.65 3965.00

Appendix 10.3. Country Allocations from JDF with No Dependence Threshold (with Compensation 50 per cent of Foreign Exchange Loss from Commodities) (Dependence Threshold 0.50)

COUNTRY

Dep. Index

Afghanistan Antigua and Barbuda Angola Bangladesh Barbados Belize Benin Bhutan Bolivia Botswana Burkina Faso Burundi Cambodia Cameroon Cape Verde CAR Chad Comoros Congo ˆ te D’Ivoire Co DRC Djibouti Dominica Equatorial Guinea Eritrea

0.56 0.04 0.01 0.14 0.30 0.92 0.48 0.18 0.64 0.14 0.52 0.93 0.26 0.67 0.27 0.49 0.61 0.59 0.12 0.71 0.91 0.28 0.55 0.31 0.32

1995

1996

1997

1998

1999

2000

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

8.93

216.35

28.78

8.04 1.27 386.23 4.63

Dep. Index 0.82 0.07 0.01 0.13 0.37 0.90 0.36 0.19 0.68 0.12 0.50 0.93 0.25 0.64 0.25 0.42 0.55 0.27 0.07 0.67 0.78 0.20 0.48 0.26 0.18

11.07

187.16

25.69 18.51

6.49

495.49 20.79

Dep. Index 0.60 0.06 0.01 0.12 0.35 0.84 0.51 0.15 0.74 0.10 0.62 1.00 0.17 0.54 0.24 0.44 0.64 0.59 0.10 0.62 0.76 0.17 0.53 0.13 0.06

9.37

201.50

65.20

8.06 1.18 510.25 28.77

Dep. Index 0.48 0.06 0.01 0.12 0.32 0.89 0.47 0.16 0.68 0.15 0.67 1.00 0.10 0.57 0.23 0.66 0.56 0.74 0.10 0.64 0.88 0.30 0.38 0.14 0.12

6.53

205.01

57.03 16.16 57.44 12.54 0.84 522.14 64.43

Dep. Index 0.70 0.05 0.01 0.11 0.30 0.96 0.32 0.15 0.67 0.11 0.53 1.00 0.08 0.59 0.22 0.58 0.46 0.57 0.08 0.58 0.68 0.69 0.43 0.14 0.32

7.42

6.98

201.60

78.92 113.85 66.16 0.87 474.49 29.72 3.87

Dep. Index 0.53 0.07 0.01 0.11 0.29 0.88 0.35 0.12 0.64 0.12 0.62 0.78 0.06 0.52 0.42 0.60 0.51 0.41 0.05 0.57 0.64 0.62 0.44 0.08 0.39

8.18

8.25

211.73

80.14

89.81 13.00

419.23 24.83

Ethiopia Fiji Gabon Gambia Ghana Grenada Guinea Guinea-Bissau Guyana Haiti Honduras Jamaica Kenya Kiribati Lao PDR Lesotho Liberia Madagascar Malawi Maldives Mali Mauritania Mauritius Mozambique Myanmar Nepal Nicaragua Niger Papua New Guinea Rwanda Samoa Sao Tome and Principe Senegal Seychelles Sierra Leone

0.97 0.51 0.12 0.92 0.49 0.80 0.74 0.92 0.66 0.34 0.50 0.71 0.66 0.95 0.47 0.08 0.06 0.86 0.89 0.74 0.61 0.59 0.30 0.81 0.91 0.17 0.82 0.72 0.61 0.74 0.86 0.73 0.44 0.46 0.77

23.18

9.10 0.46

25.51 57.71 2.95 6.75

20.88 112.79 22.58 45.54

0.30

0.95 0.44 0.10 0.91 0.69 0.81 0.64 0.91 0.71 0.34 0.45 0.74 0.61 0.96 0.38 0.05 0.08 0.95 0.80 0.83 0.73 0.61 0.30 0.62 0.91 0.17 0.92 0.58 0.58 0.84 0.91 0.88 0.46 0.30 0.52

31.22

95.32

10.12

0.13

14.77 51.99 2.45 7.43

40.98 158.55 50.16 55.30 31.54 0.20 0.26

0.90 0.40 0.12 1.00 0.58 0.79 0.61 0.57 0.55 0.28 0.44 0.77 0.62 0.95 0.19 0.05 0.11 0.88 0.94 0.77 0.53 0.65 0.29 0.74 0.73 0.34 0.83 0.66 0.56 0.65 0.93 0.66 0.47 0.63 0.58

35.02

76.98

8.83

0.27

15.13 51.67 4.35 13.46

51.11 203.86 63.36 59.63 55.47 0.64 0.36 5.49

0.95 0.41 0.20 0.57 0.47 0.79 0.52 0.71 0.72 0.20 0.53 0.76 0.73 0.91 0.15 0.04 0.12 0.88 0.72 0.76 0.56 0.52 0.28 0.60 0.61 0.15 0.62 0.74 0.59 0.84 0.94 0.50 0.46 0.74 0.34

59.57

12.95 0.12

0.26

16.40 85.75 4.35 7.79

43.20 241.97 42.62 79.38 42.24 0.71 0.39 6.93

0.90 0.37 0.20 0.99 0.54 0.57 0.53 0.77 0.63 0.13 0.42 0.81 0.64 0.81 0.20 0.04 0.16 0.92 0.77 0.61 0.43 0.52 0.26 0.51 0.58 0.13 0.74 0.85 0.52 0.80 0.92 0.74 0.45 0.76 0.32

78.16

73.69 1.11 23.51 18.16

43.76 0.27

16.45 58.47 3.16

44.71 267.74 65.98 84.26 33.15 0.62 0.72 8.94

0.84 0.42 0.15 0.96 0.68 0.49 0.51 0.69 0.70 0.16 0.55 0.81 0.64 0.68 0.19 0.03 0.29 0.53 0.82 0.54 0.47 0.60 0.25 0.48 0.41 0.08 0.72 0.83 0.51 0.80 0.91 0.95 0.48 0.65 0.31

112.00

35.87 20.00 25.44 50.58 0.21

12.57 67.08 3.72

78.96 98.07 26.59 0.32 1.12 10.48

(Continued )

Appendix 10.3. (Continued )

COUNTRY

Dep. Index

Solomon Islands Somalia St Kitts and Nevis St Lucia St Vincent Sudan Suriname Swaziland Togo Tonga Trinidad and Tobago Tuvalu Uganda United Rep. of Tanzania Vanuatu Vietnam Yemen Zambia

0.98 0.59 0.86 0.52 0.99 0.91 0.96 0.34 0.52 0.81 0.08 0.48 0.89 0.69 0.73 0.38 0.05 0.80

TOTAL

1995

1996

1997

1998

1999

2000

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

5.30

27.84 26.60 24.70

125.08 100.73

1292.22

Dep. Index 0.97 0.56 0.61 0.72 0.80 0.92 0.95 0.42 0.50 0.81 0.08 0.32 0.84 0.69 0.79 0.33 0.03 0.77

4.14 6.29

42.44 22.74 34.13 0.21

218.81 115.02 0.51

56.03 1815.97

Dep. Index 0.99 0.70 0.76 0.61 0.76 0.96 0.79 0.44 0.51 0.92 0.10 0.71 0.83 0.78 0.86 0.32 0.03 0.96

9.85 5.52

37.93 20.17 28.60 0.46 0.02 219.51 78.60 1.40

72.17 1944.22

Dep. Index 1.00 0.48 0.57 0.62 0.85 0.86 0.92 0.49 0.62 0.88 0.11 0.67 0.85 0.89 1.00 0.35 0.06 0.56

6.40

56.82 22.25 18.68 0.27 0.01 209.05 137.44 1.36

113.58 2152.61

Dep. Index 0.94 0.62 0.45 0.68 0.78 0.46 0.88 0.45 0.49 0.88 0.09 0.25 0.90 0.69 0.95 0.30 0.03 0.57

3.44 6.50 1.66 1.64 30.56

0.36

257.85 102.38 0.81

179.95 2391.86

Dep. Index 0.91 0.63 0.39 0.84 0.81 0.21 0.91 0.49 0.48 0.87 0.06 0.49 0.91 0.58 0.91 0.26 0.03 0.69

2.51 9.86 1.84 1.84 32.59

0.28

367.27 114.94 0.57

123.65 2053.54

Appendix 10.4. Country Allocations from JDF with No Dependence Threshold (with Compensation 50 per cent of Foreign Exchange Loss from Commodities) (Dependence Threshold 0.75)

COUNTRY

Dep. Index

Afghanistan Antigua and Barbuda Angola Bangladesh Barbados Belize Benin Bhutan Bolivia Botswana Burkina Faso Burundi Cambodia Cameroon Cape Verde CAR Chad Comoros Congo ˆ te D’Ivoire Co DRC Djibouti Dominica

0.56 0.04 0.01 0.14 0.30 0.92 0.48 0.18 0.64 0.14 0.52 0.93 0.26 0.67 0.27 0.49 0.61 0.59 0.12 0.71 0.91 0.28 0.55

1995

1996

1997

1998

1999

2000

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

28.78

4.63

Dep. Index 0.82 0.07 0.01 0.13 0.37 0.90 0.36 0.19 0.68 0.12 0.50 0.93 0.25 0.64 0.25 0.42 0.55 0.27 0.07 0.67 0.78 0.20 0.48

11.07

25.69

20.79

Dep. Index 0.60 0.06 0.01 0.12 0.35 0.84 0.51 0.15 0.74 0.10 0.62 1.00 0.17 0.54 0.24 0.44 0.64 0.59 0.10 0.62 0.76 0.17 0.53

65.20

28.77

Dep. Index 0.48 0.06 0.01 0.12 0.32 0.89 0.47 0.16 0.68 0.15 0.67 1.00 0.10 0.57 0.23 0.66 0.56 0.74 0.10 0.64 0.88 0.30 0.38

57.03

64.43

Dep. Index 0.70 0.05 0.01 0.11 0.30 0.96 0.32 0.15 0.67 0.11 0.53 1.00 0.08 0.59 0.22 0.58 0.46 0.57 0.08 0.58 0.68 0.69 0.43

6.98

78.92

Dep. Index 0.53 0.07 0.01 0.11 0.29 0.88 0.35 0.12 0.64 0.12 0.62 0.78 0.06 0.52 0.42 0.60 0.51 0.41 0.05 0.57 0.64 0.62 0.44

8.25

80.14

(Continued )

Appendix 10.4. (Continued )

COUNTRY

Dep. Index

Equatorial Guinea Eritrea Ethiopia Fiji Gabon Gambia Ghana Grenada Guinea Guinea-Bissau Guyana Haiti Honduras Jamaica Kenya Kiribati Lao PDR Lesotho Liberia Madagascar Malawi Maldives Mali Mauritania Mauritius Mozambique

0.31 0.32 0.97 0.51 0.12 0.92 0.49 0.80 0.74 0.92 0.66 0.34 0.50 0.71 0.66 0.95 0.47 0.08 0.06 0.86 0.89 0.74 0.61 0.59 0.30 0.81

1995

1996

1997

1998

1999

2000

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

Foreign Exchange Loss from Commodities (1984–86 prices)

23.18

9.10

25.51 57.71

20.88

Dep. Index 0.26 0.18 0.95 0.44 0.10 0.91 0.69 0.81 0.64 0.91 0.71 0.34 0.45 0.74 0.61 0.96 0.38 0.05 0.08 0.95 0.80 0.83 0.73 0.61 0.30 0.62

31.22

10.12

0.13

14.77 51.99 2.45

Dep. Index 0.13 0.06 0.90 0.40 0.12 1.00 0.58 0.79 0.61 0.57 0.55 0.28 0.44 0.77 0.62 0.95 0.19 0.05 0.11 0.88 0.94 0.77 0.53 0.65 0.29 0.74

35.02

0.27

15.13 51.67 4.35

Dep. Index 0.14 0.12 0.95 0.41 0.20 0.57 0.47 0.79 0.52 0.71 0.72 0.20 0.53 0.76 0.73 0.91 0.15 0.04 0.12 0.88 0.72 0.76 0.56 0.52 0.28 0.60

59.57

0.26

16.40 4.35

Dep. Index 0.14 0.32 0.90 0.37 0.20 0.99 0.54 0.57 0.53 0.77 0.63 0.13 0.42 0.81 0.64 0.81 0.20 0.04 0.16 0.92 0.77 0.61 0.43 0.52 0.26 0.51

78.16

23.51

43.76 0.27

16.45 58.47

Dep. Index 0.08 0.39 0.84 0.42 0.15 0.96 0.68 0.49 0.51 0.69 0.70 0.16 0.55 0.81 0.64 0.68 0.19 0.03 0.29 0.53 0.82 0.54 0.47 0.60 0.25 0.48

50.58

67.08

Myanmar Nepal Nicaragua Niger Papua New Guinea Rwanda Samoa Sao Tome and Principe Senegal Seychelles Sierra Leone Solomon Islands Somalia St Kitts and Nevis St Lucia St Vincent Sudan Suriname Swaziland Togo Tonga Trinidad and Tobago Tuvalu Uganda United Rep. of Tanzania Vanuatu Vietnam Yemen Zambia TOTAL

0.91 0.17 0.82 0.72 0.61 0.74 0.86 0.73 0.44 0.46 0.77 0.98 0.59 0.86 0.52 0.99 0.91 0.96 0.34 0.52 0.81 0.08 0.48 0.89 0.69 0.73 0.38 0.05 0.80

112.79 22.58 45.54

27.84 26.60

125.08

530.23

0.91 0.17 0.92 0.58 0.58 0.84 0.91 0.88 0.46 0.30 0.52 0.97 0.56 0.61 0.72 0.80 0.92 0.95 0.42 0.50 0.81 0.08 0.32 0.84 0.69 0.79 0.33 0.03 0.77

158.55 50.16

0.20 0.26

4.14

42.44 22.74

0.21

218.81 0.51

56.03 722.30

0.73 0.34 0.83 0.66 0.56 0.65 0.93 0.66 0.47 0.63 0.58 0.99 0.70 0.76 0.61 0.76 0.96 0.79 0.44 0.51 0.92 0.10 0.71 0.83 0.78 0.86 0.32 0.03 0.96

63.36

0.64

9.85

37.93 20.17

0.46

219.51 78.60 1.40

72.17 704.52

0.61 0.15 0.62 0.74 0.59 0.84 0.94 0.50 0.46 0.74 0.34 1.00 0.48 0.57 0.62 0.85 0.86 0.92 0.49 0.62 0.88 0.11 0.67 0.85 0.89 1.00 0.35 0.06 0.56

0.71

6.40

56.82 22.25

0.27

209.05 137.44 1.36

636.34

0.58 0.13 0.74 0.85 0.52 0.80 0.92 0.74 0.45 0.76 0.32 0.94 0.62 0.45 0.68 0.78 0.46 0.88 0.45 0.49 0.88 0.09 0.25 0.90 0.69 0.95 0.30 0.03 0.57

84.26

0.62

8.94 3.44

1.64 30.56

0.36

257.85 0.81

694.99

0.41 0.08 0.72 0.83 0.51 0.80 0.91 0.95 0.48 0.65 0.31 0.91 0.63 0.39 0.84 0.81 0.21 0.91 0.49 0.48 0.87 0.06 0.49 0.91 0.58 0.91 0.26 0.03 0.69

98.07

0.32 1.12

2.51

1.84 1.84 32.59

0.28

367.27 0.57

712.46

This page intentionally left blank

References

Abrego, L. and Ross, D. C. (2002), ‘Debt Relief Under HIPC Initiative: Context and Outlook for Debt Sustainability and Resource Flows’, WIDER Discussion Paper No. 2002/44, United Nations University. Addison, T., McGillivray, M., and Odedokun, M. (2003), ‘Donor Funding of Multilateral Aid Agencies—Determining Factors and Revealed Burden Sharing’, WIDER Discussion Paper No. 2003/17, United Nations University. Agenor P. R. and Montiel P. J. (1999), Development Macroeconomics 2nd edn., Princeton: Princeton University Press. Aiello, F. (1999), ‘The Stabilisation of LDC’s Export Earnings: The Impact of EU STABEX Programme’, International Review of Applied Economics, 13 (1). Alauddin, M. and Tisdell, C. (1995), ‘Labor Absorption and Agricultural Development: Bangladesh’s Experience and Predicament’, World Development 23 (2), 281–98. —— Headey, D., and Rao, D. S. P. (2005), ‘Explaining Agricultural Productivity Levels and Growth: An International Perspective’, Unpublished paper, University of Queensland, Brisbane. Amin, A. A., Douya, E., and Mbeaoh, A. (2002), ‘The Impact of Trade Liberalization and Domestic PolicyReformsonthe AgriculturalSectorinCameroon’, ch. 6 inD.Bigman(ed.), Globalization and the Developing Countries: Emerging Strategies for Rural Development and Poverty Alleviation, International Service for National Agricultural Research, The Hague. Amjadi, A. and Yeats, A. (1995), ‘Have Transport Costs Contributed to the Relative Decline of African Exports? Some Preliminary Evidence’, World Bank Policy Research Working Paper, Washington, D.C.: World Bank. —— Reincke, U., and Yeats, A. (1996), ‘Did External Barriers Cause the Marginalisation of Sub-Saharan Africa in World Trade?’, World Bank Policy Research Working Paper, no. 1586, Washington, D.C.: World Bank. Andersen, R. W. and Gilbert, C. L. (1988). ‘Commodity Agreements and Commodity Markets—Lessons from Tin’, Economic Journal, 98, (1), 1–15. APEP (2005), Commodities: coffee, Uganda Agricultural Productivity Enhancement Program, Kampala [www.apepuganda.org/htm/commodities_coffee.htm]. Appleyard, D. R. and Field, A. J. (1998), ‘International Economics, 3rd edn’, New York: McGraw-Hill. Ardeni, P. G. and Wright, B. (1992), ‘The Prebisch-Singer Hypothesis: a Reappraisal Independent of Stationarity Hypothesis’ Economic Journal, 102, 803–12. Arndt, C. (2005), ‘The Doha Trade Round and Mozambique’, ch. 5 in T. W. Hertel and L. A. Winters (eds.), Poverty impacts of a WTO agreement: Putting development back into the Doha Agenda [http://www.worldbank.org/trade/wto].

329

References Athukorala, P. C. (2000), ‘Manufactured Exports and Terms of Trade of Developing Countries: Evidence from Sri Lanka’, Journal of Development Studies, 36, (5). —— and Riedel, J. (1996), ‘Modelling NIE Exports: Aggregation, Quantitative Restrictions and Choice of Econometric Methodology’, Journal of Development Studies, 33 (1). 81–98. Atkins, J. P., Mazzi, S., and Easler, C. D. (2000), ‘A Commonwealth Vulnerability Index for Developing Countries: The Position of Small States’, Commonwealth Secretariat, London. Balat, J. F. and Porto, G. G. (2005), ‘The WTO Doha Round, cotton sector dynamics and poverty trends in Zambia’, ch. 6 in T. W. Hertel and L. A. Winters (eds.), Poverty impacts of a WTO agreement: Putting development back into the Doha Agenda. [http://www.worldbank.org/trade/wto]. Banerjee, A., Dolado, J. J., Galbraith, J. W., and Hendry, D. F. (1993), Cointegration, Error Correction, and the Econometric Analysis of Non-stationary Data, Oxford University Press. Beckerman W. and Jenkinson T. (1986), ‘What Stopped the Inﬂation: Unemployment or Commodity Prices?’, Economic Journal, 96, 39–54. Berlage, L., Cassimon, D., Dreze, J., and Reding, P. (2003), ‘Prospective Aid and Indebtedness Relief: A Proposal’, World Development, 31 (10), 1635–54. Bernall, R. (2001), ‘Globalisation and Small Developing Economies: Challenges and opportunities’, in Peretz, D., Faruqi, R., and Kisanga, E. (eds.) Small States in the Global Economy, Commonwealth Secretariat and World Bank. Berthelot, P. (2001), ‘Small States in the Lome´ Convention and in the New ACP-EU Partnership Agreement: A Commentary’, Small States: Economic Review and Basic Statistics, 6, Commonwealth Secretariat, London. Bevan, D. L., Collier, P., and Gunning, J. W. (1993), ‘Trade Shocks in Developing Countries: Consequences and Policy Responses’, European Economic Review, 37, 557–65. Beveridge, S. and Nelson, C. R. (1981), ‘A New Approach to Decomposition of Economic Time Series into Permanent and Transitory Components with Particular Attention to Measurement of Business Cycle’, Journal of Monetary Economics, 7, 151–74. Bhagwati, J. (1958), ‘Immiserizing growth: a geometrical note’, Quarterly Journal of Economics 25, 201–5. Bigman, D. (2002), ‘Diversifying Agricultural Exports and Production in Agriculture’, ch. 11 in D. Bigman (ed.), Globalization and the Developing Countries: Emerging Strategies for Rural Development and Poverty Alleviation, International Service for National Agricultural Research, The Hague. Birdsall, N. (2002), ‘Beyond HIPC: Secure Sustainable Debt Relief for Poor Countries’, Paper prepared for the Commonwealth Finance Ministers Meeting 2002, mimeo. Birdsall, N. and Hamoudi, A. (2002), ‘Commodity Dependence, Trade and Growth: When Openness is Not Enough’, Centre for Global Development Working Paper No. 7. Birdsall, N. and Williamson, J. with Deese, B. (2002), ‘Delivering on Debt Relief: From IMF Gold to a New Aid Architecture’, Centre for Global Development and the Institute for International Economics, Washington, D.C. —— Claessens, S., and Diwan, I. (2002), ‘Will HIPC Matter? The Debt Game and Donor Behaviour in Africa’, WIDER Discussion Paper No. 2002/50, United Nations University.

330

References Bleaney M. F. (1993), ‘Manufactured Exports of Developing Countries and their Terms of Trade since 1965: A Comment’, World Development, 21 (10), 1615–16. Bleaney, M. F. and Greenaway, D. (1993), ‘Long-run Trends in the Relative Price of Primary Commodities and in the Terms of Trade of Less Developed Countries’, Oxford Economic Papers, 45, 349–63. Bloch H. and Sapsford D., (1992), ‘Post War Movements in Prices of Primary Products and Manufactured Goods’, Journal of Post Keynesian Economics, 14, 249–67. —— (1997), ‘Some estimates of Prebisch and Singer Effects on the Terms of Trade between Primary Producers and Manufacturers’, World Development, 25, (11). —— (2000), ‘Whither Terms of Trade? An Elaboration of the Prebisch-Singer Hypothesis’, Cambridge Journal of Economics, 24, 461–81. Bonaglia, F. and Fukasaku, K. (2003), ‘Export Diversiﬁcation in Low Income Countries: An International Challenge After Doha’, Technical Papers No. 9, OECD Development Centre. Borensztein, E. and Reinhart, C. M. (1994), ‘The Macroeconomic Determinants of Real Commodity Prices’, IMF Staff Papers, 41, 236–61. Borensztein, E., Khan, M. S., Reinhart, C. M., and Wickham, P. (1994), ‘The Behaviour of Non-Oil Commodity Prices’, IMF Occasional Paper 112, Washington, DC. Brun, J.-F., Carrere, C., Guillaumont, P., and de Melo, J. (2002), ‘Has Distance Died? No and Why’, mimeo. Bureau, J.-C. and Gohin, A. (2005), Assessing the Effects of Sugar Market Liberalization: Impact on the Global Sugar Markets, Contributed Paper to the 11th Congress of the European Association of Agricultural Economists, Copenhagen, 24–7 August. Cashin, P. and McDermott, C. J. (2002), ‘The Long-Run Behaviour of Commodity Prices: Small Trends and Big Variability’, IMF Staff Papers, 49 (2), 175–99. —— Liang, H., and McDermott, C. J. (1999). ‘How Persistent Are Shocks to World Commodity Prices?’, IMF Working Paper, Washington, DC. Cassen, R. (1986), ‘Does Aid Work?’, Oxford: Clarendon Press. —— and Associates (1994), ‘Does Aid Work?: Report to an Intergovernmental Task Force’, Oxford: Clarendon Press. Chang, C., Eduardo, F., and Serven, L. (1999), ‘Measuring Aid Flows—A New Approach’, World Bank Policy Research Work Paper No. 2050. Washington, D.C: World Bank. Charemza, W. W. and Deadman, D. F. (1992), ‘New Directions in Econometric Practice’, Edward Elgar: Cheltenham. Chen and Stocker (1997), ‘Composite Commodities and the Prebisch-Singer Thesis’, World Development, vol. 25, no. 11. Choraria, J. (2005), ‘A note on Commodity Value Chains Compression—Coffee, Cocoa and Sugar’, Unpublished paper, Commonwealth Secretariat, London. Coelli, T. and Rao, D. S. P. (2004), ‘Total Factor Productivity Growth in Agriculture: a Malmquist Index Analysis of 93 countries, 1980–2000’, in D. Colman and N. Vink (eds), Reshaping Agriculture’s Contribution to Society: Proceedings of the 25th International Conference of Agricultural Economists Held at Durban, South Africa, 16–22 August, 2003, Blackwell, Oxford. Cohen, D. (2000), ‘The HIPC Initiative: True and False Promises’, OECD Working Paper, Paris: OECD Development Centre.

331

References Collier, P. (2002), ‘Primary Commodity Dependence and Africa’s Future’, mimeo: World Bank: Washington, D.C. —— and Dehn, J. (2002), ‘Aid, Shocks, and Growth’, Policy Research Working Paper 2443, World Bank: Washington, D.C. —— and Dollar, D. (2001), ‘Aid, Risk and the Special Concerns of Small States’, in Peretz, D., Faruqi, R., and Kisanga, E. (eds.), Small States in the World Economy, Commonwealth Secretariat and World Bank. —— and Gunning, J. W. (1999), ‘Explaining African Economic Performance’, Journal of Economic Literature, 37, 64–111. Commission of the European Communities (2004), ‘Agricultural Commodity Chains, Dependence and Poverty—A Proposal for an EU Action Plan’, Communication from the Commission to the Council and the European Parliament, Brussels. Common Fund for Commodities (2002). ‘Common Fund for Commodities Annual Report, 2001’, Amsterdam: Common Fund for Commodities. Commonwealth Secretariat (1997) ‘A Future for Small States: Overcoming Vulnerability’, Report by a Commonwealth Advisory Group. London: Commonwealth Secretariat. Cuddington, J. T. (1992) ‘Long Run Trends in 26 Primary Commodity Prices: A disaggregated look at the Prebisch-Singer Hypothesis’ Journal of Development Economics, 39, 207–27. —— and Urzua, C. M. (1989), ‘Trends and Cycles in the Net Barter Terms of Trade: A New Approach’, Economic Journal, 99, 426–32. Cypher, J. M. and Dietz, J. L. (2004), ‘The Process of Economic Development’, 2nd edn., Routledge, New York. Danielson, A. and Mjema, G. (2001), ‘Tanzania 2000: Growth, Multilateral Debt Relief and Programme Aid’, SIDA Country Economic Report, 2001/3, Stockholm: SIDA. Daseking, C. and Powell, R. (1999), ‘From Toronto Terms to the HIPC Initiative: A Brief History of Debt Relief for Low Income Countries’, IMF Working Paper, WP/99/142. Deaton, A. (1999), ‘Commodity Prices and Growth in Africa’, Journal of Economic Perspectives, 13, no. 3. —— and Laroque, G. (1992), ‘On the Behaviour of Commodity Prices’, Review of Economic Studies, 59, 1–24. —— and Miller R. (1996), ‘International Commodity Prices, Macroeconomic Performance and Politics in Sub-Saharan Africa’, Journal of African Economies, 5(3), Supplement. —— (2003), ‘A Model of Commodity Prices after Sir Arthur Lewis’, Journal of Development Economics, 71, 289–310. DEFRA (2004), ‘Investigation of the Determinants of Farm-Retail Price Spreads’ by London Economics. Dehn, J. (2001), ‘Commodity Prices Uncertainty and Shocks: Implications for Economic Growth’, paper presented at the Royal Economic Society Annual Conference 2001, University of Durham, 9–11 April 2001. —— Gilbert, C., and Varangis, P. (2003), ‘Commodity Price Volatility’, mimeo, World Bank: Washington, D.C. Devarajan, S., Easterly, W., and Howard, P. (2002), ‘Low Investment is not the Constraint on African Development’, Centre for Global Development Discussion Paper. Devarajan, S. and Swaroop, V. (1998), ‘The Implications of Foreign Aid Fungibility for Development Assistance’, World Bank Working Papers 2022, World Bank: Washington, D.C.

332

References Diakosavvas D. and Scandizzo P. L., (1991), ‘Trends in the Terms of Trade of Primary Commodities, 1900–1982: The Controversy and Its Origins’, Economic Development and Cultural Change, 39 (2), 231–64. Dickey, D. A. and Fuller, W. A. (1979), ‘Distribution of the Estimators for Autoregressive Time Series with a Unit Root’, Journal of the American Statistical Association, 74, 427–31. —— (1981), ‘Likelihood Ratio Statistics for Autoregressive Time Series With A Unit Root’, Econometrica, 49, 1057–72. Dodhia, D. (2002), ‘Progress Report on Implementing the HIPC Initiative’, Paper Prepared for the Commonwealth Finance Ministers’ Meeting 2002, mimeo: Commonwealth Secretariat, London. Doornik, J. A. and Hendry, D. (1997), ‘Modelling Dynamic Systems Using PcFiml 9 for Windows’, London: Timberlake Consulting. Duncan, R. C. (1994), ‘A Fresh Look at Some Long-Held Beliefs about Primary Commodities’, Paciﬁc Economic Bulletin 9 (1), 53–57. Easterly, W. (2001), ‘Growth Implosions and Debt Explosions: Do Growth Slowdowns Cause Public Debt Crisis?’ Contributions to Macroeconomics, 1, (1), 1–24. —— (2002), ‘How Did Heavily Indebted Poor Countries Become Heavily Indebted? Reviewing Two Decades of Debt Relief’, World Development, 30, 1677–96. —— and Levine, R. (2002), ‘Tropics, Germs and Crops: How Endowments Inﬂuence Economic Development’, Centre for Global Development Working Paper. Engle, R. F. and Granger, C. W. J. (1987), ‘Cointegration and Error Correction: Representation, Estimation and Testing’, Econometrica, 55, 251–76. Eurodad (2001), ‘Rethinking HIPC Debt Sustainability’ accessed from www.eurodad.org on 10 July 2003. Feeny, S. and McGillivray, M. (2002), ‘Aid Public Sector Fiscal Behaviour and Developing Country Debt’, WIDER Discussion Paper No. 2002/40, United Nations University. Fitter, R. and Kaplinsky, R. (2001), ‘Can an Agricultural ‘‘Commodity’’ be De-Commodiﬁed, and if so, Who is to Gain?’, Institute of Development Studies, 2001. Frankel, J. A. and Romer, D (1999), ‘Does Trade Cause Growth?’, American Economic Review, 89, 3, 379–99. Fuglie, K.O. (2004), ‘Productivity Growth in Indonesian Agriculture, 1961–2000’, Bulletin of Indonesian Economic Studies 40 (2), 209–25. Fulginiti, L. E. and Perrin, R. K. (1997), ‘LDC agriculture: non-parametric Malmquist productivity indexes’, Journal of Development Economics 53 (2), 373–90. Gallup, J. L. and Sachs, J. D. (2000), ‘Agriculture, Climate, and Technology: Why are the Tropics Falling Behind?’, American Journal of Agricultural Economics 82, 731–7. Gautam, M. (2003), ‘Debt Relief for the Poorest: An OED Review of the HIPC Initiative’, World Bank Operations Evaluation Department, World Bank, Washington, D.C. Geda, A. (2002), ‘Debt Issues in Africa: Thinking Beyond the HIPC Initiative to Solving Structural Problems’, WIDER Discussion Paper No. 2002/35, United Nations University. Gerefﬁ, G. (1996) ‘Global Commodity Chains: New Forms of Coordination and Control Among Nations and Firms in International Industries’, Competition and Change, 1(4), 427–39.

333

References Gerefﬁ, G. and Korzeniewicz, M. (eds.) (1994), Commodity Chains and Global Capitalism, Westport, CT: Praeger. Gibbon, P. (2001), ‘Upgrading primary production: a global commodity chain approach’, World Development, 29 (2), 345–63. Gilbert, C. L. (1987), ‘International Commodity Agreements: Design and Performance’, World Development, 15 (5), 591–616. —— (1989) ‘The Impact of Exchange Rates and Developing Country Debt on Commodity Prices’, Economic Journal, 99, 773–84. —— (1996). ‘International Commodity Agreements: An Obituary Notice’, World Development, 24 (1), 1–19. Gilbert, R. and Tabova, A. (2003), ‘Realignment of Debt Service Obligations and Ability to Pay in Concessional Lending: Feasibility and Modalities’, mimeo. Godfrey, L. G. (1978), ‘Testing for Higher Order Serial Correlation in Regression Equations When the Regressors Include Lagged Dependent Variables’, Econometrica, 46, 1293–1301. Granger, C. J. W. and Newbold, P. (1974), ‘Spurious Regression in Economics’, Journal of Econometrics, 2, 111–20. Greenaway and Morgan (1999), ‘The Economics of Commodity Markets’, Aldershot: Edward Elgar. Grilli, E. R. and Yang, M. C. (1988), ‘Primary Commodity Prices, Manufactured Goods Prices, and Terms of Trade of Developing Countries: What the Long-run Shows’, World Bank Economic Review, 1–47. Grynberg, R. (2001), ‘Trade Policy Implications for Small Vulnerable States of the Global Trade Regime Shift’, in Pertez, D., Faruqi, R., and Kisanga, E. J. (eds.), Small States in the Global Economy, Commonwealth Secretariat and World Bank, 267–328. Grynberg, R. and Razzaque, M. A. (2003), ‘Trade Problems of Small States’, Economic Affairs Division Working Paper, Commonwealth Secretariat, London. —— (2004), ‘The Trade Performance of Small States’, Economic Paper 59, Commonwealth Secretariat, London. Grynberg, R. and Silva, S. (2004), ‘Preference-Dependent Economies and Multilateral Liberalization: Impacts and Options’, Commonwealth Secretariat Report, Commonwealth Secretariat, London. Gujarati, D. N. (1995), ‘Basic Econometrics’, Sydney: McGrawHill. —— (2003), ‘Basic Econometrics’, Sydney: McGrawHill. Hadass, Y. and Williamson, J. (2002), ‘Terms of Trade Shocks and Economic performance 1870–1940: Prebisch and Singer Revisited’, NBER Working Paper No. W8188, March 2001. Hall, S. (1986), ‘An Application of the Granger and Engle Two-Step Estimation Procedure to United Kingdom Aggregate Wage Data’, Oxford Bulletin of Economics and Statistics, 48 (3), 229–40. Harris, R. (1995), ‘Using Cointegration Analysis in Econometric Modelling’, Prentice Hall and Harvester Wheatsheaf. Harvey, A. C. (1989), ‘Forecasting, Structural Time Series Models and the Kalman Filter. Cambridge: Cambridge University Press. Healy, S. (2004), ‘La baisse des prix des produits agricoles: quelles consequences pour les ˆ te d’Ivoire et du riz en Guine´e Conakry’, CERDI. pays africains? Cas du cacao en Co

334

References Henao, J. and Baanante, C. (1999), ‘Nutrient Depletion in the Agricultural Soils of Africa’, 2020 Brief No. 62, International Food Policy Research Institute, Washington, DC. Hendry, D. F. (1980), ‘Econometrics—Alchemy or Science’, Economica 47, 387–406. —— (1986), ‘Econometric Modelling with Cointegrated Variables: An Overview’, Oxford Bulletin of Economics and Statistics, 48, 201–12. —— (1995), ‘Dynamic Econometrics’, Oxford University Press, 577–619. —— (1999), ‘An Econometric Analysis of US Food Expenditure, 1931–1989’, ch. 17 in Magnus, J. R. and Morgan, M. S. (eds.), Methodology and Tacit Knowledge: Two Experiments in Econometrics, John Wiley and Sons. Herrmann, R. and Weiss D., (1995), ‘A Welfare Analysis of the EC-ACP Sugar Protocol’, Journal of Development Studies, 31, 6. —— Burger, K., and Smit, H. P. (1993). International Commodity Policy: A Quantitative Analysis, London: Routledge. Hertel, T. W. and Winters, L. A. (2005), ‘Poverty impacts of a WTO agreement: synthesis and overview’, ch. 1 in T. W. Hertel and L. A. Winters (eds), Poverty impacts of a WTO agreement: Putting development back into the Doha Agenda. [http://www.worldbank. org/trade/wto] Hewitt, A. (1987), ‘STABEX and Commodity Export Compensation Schemes: Prospects for Globalization’, World Development, 15, 5, 617–31. —— (1993), ‘Economic Crisis in Developing Countries: New Perspectives on Commodities, Trade and Finance—Essays in Honour of Alfred Maizels’ in Nissanke M., and Hewitt, A. (eds.), London: Pinter. —— and Gillson, I. (2003), ‘A Review of the Trade and Poverty Content in PRSPs and Loan-Related Documents’, Overseas Development Institute. Hjertholm, P. (1999), ‘Analytical History of Heavily Indebted Poor Country (HIPC) Debt Sustainability Targets’, paper presented at the World Bank/Nordic Working Seminar: Review of the HIPC Initiative, Oslo, on 4 March 1999. Hoekman, B., and Braga, C.P. (1997), ‘Protection and Trade in Services: A Survey’, Open Economies Review 8 (3): 285–308. —— Ng, F., and Olarreaga, M. (2001). ‘Tariff Peaks in the Quad and Least Developed Country Exports’, CEPR Discussion Paper No. 2747, Centre for Economic Policy Research, London. Holden, D. and Perman, R. (1994), ‘Unit Roots and Cointegration for the Economists’, in B. Bhaskara Rao (ed.) Cointegration for the Applied Economists, St Martin’s Press. Hopkins, T. K. and Wallerstein, I. (1986), ‘Commodity Chains in the World-Economy’, Review, 10 (1), 157–70. IAC (2004), ‘Realizing the Promise and Potential of African Agriculture: Science and Technology Strategies for Improving Agricultural Productivity and Food Security in Africa’, Report to the United Nations, InterAcademy Council. IADB (2003), ‘Productivity and Competitiveness in the Jamaican Economy’, Economic and Sector Study Series RE3–03–002, Inter-American Development Bank, Washington, DC. Ianchoviche, E., Mattoo, A., and Olarreaga, M. (2001), ‘Unrestricted Market Access for Sub-Saharan Africa: How Much Is It Worth and Who Pays?’ CEPR Discussion Paper No. 2820, Centre for Economic Policy Research, London.

335

References International Coffee Organization (2003), ‘Data Concepts and Variables used in the Statistics of the Organization’, International Coffee Organization International Cotton Advisory Committee (2002), ‘Cotton: Review of the World Situation’, vol. 55, no. 5. International Development Association (IDA) (2003), ‘Heavily Indebted Poor Countries Initiative—Statistical Update’, April 11, accessed from www.worldbank.org/hipc on 22 August 2003. International Monetary Fund (IMF) (1994), ‘Adjustment, not Resistance, the Key to Dealing with Low Commodity Prices’, IMF Survey 23, 350–52. —— (2000), ‘International Financial Statistics Yearbook’, IMF, Washington, DC. —— and International Development Association (IDA) (2002), ‘The Enhanced HIPC Initiative and the Achievement of the Long-run External Debt Sustainability’, April 15, accessed from www.worldbank.org/hipc on 22 August 2003. —— (2004), ‘International Financial Statistics (and previous issues)’, International Monetary Fund, Washington, DC. International Task Force on Commodity Risk Management in Developing Countries (1999), ‘Dealing with Commodity Price Volatility in Developing Countries: A Proposal for a Market-Based Approach’, Washington, D.C: World Bank. Jarque, C. M. and Bera A. K. (1980), ‘Efﬁcient Tests for Normality, Homoscedasiticy and Serial Independence of Regression Residuals’, Economics Letters, 6, 255–9. Johansen, S. and Juselius, K. (1990), ‘Maximum Likelihood Estimation and Inference on Cointegration With Applications to the Demand for Money’, Oxford Bulletin of Economics and Statistics, 52, 169–210. Johnson, D. K. N. and Evenson, R. E. (2000), ‘How far away is Africa? Technological Spillovers to Agriculture and Productivity’, American Journal of Agricultural Economics 82, 743–9. Kanbur, R., Sandler, T., and Morrison, K. (1999), ‘The Future of Development Assistance: Common Pools and International Public Goods’, ODC Policy Essay No 25. Kandiero, T. and Randa, J. (2004), ‘Agricultural Exports: Important Issues for SubSaharan Agriculture’, African Development Review 16 (1), 1–35. Kellard N. and Wohar M. E. (2002), ‘Trends and Persistence in Primary Commodity Prices’, mimeo. Khatkhate, D. R. (1966), ‘Debt Servicing as an Aid to Promotion of Trade of Developing Countries’, Oxford Economic Papers. Kiringai, J. (2002), ‘Debt and PRSP Conditionality: The Kenya Case’, WIDER Discussion Paper No. 2002/54, United Nations University. Knapp, R. (2004), ‘Sugar and the European Union: implication of WTO ﬁndings, and reform’, Horticultural and Tropical Products Division, Foreign Agricultural Service, USDA, Washington, D.C. (www.fas.usda.gov/htp/sugar/2004). Kuhnen, F. (1987), ‘Causes of Underdevelopment and Concepts for Development: An Introduction to Development Theories’, Journal of Development Studies (Peshawar) 8, 11–25. Kurer, O. (2001), ‘Land Tenure and Sugar Production in Fiji: Property Rights and Economic Performance’, Paciﬁc Economic Bulletin 16 (2), 94–105. Lee, T., White, H., and Granger, C. W. J. (1993) ‘Testing for Neglected Nonlinearity in Time Series Models’, Journal of Econometrics, 56, 269–90.

336

References Leon, J. and Soto, R. (1997), ‘Structural Breaks and Long-run Trends in Commodity Prices’, Journal of International Development, 9, 347–66. Lewis, A. (1952), ‘World Production, Prices and Trade, 1870–1950, Manchester School of Economics and Social Studies, 20. —— (1954), ‘Economic Development with Unlimited Supplies of Labour’, Manchester School, vol. 22, 130–90. Lima˜o, N. and Venables, A. J. (2001), ‘Infrastructure, Geographical Disadvantage, Transport Costs and Trade’, World Bank Economic Review. Lipsey, R. E. (1994), ‘Quality Change and other Inﬂuences on Measures of Export Prices of Manufactured Goods and the Terms of Trade between Primary Products and Manufactures’, NBER Working Paper No. 4671, National Bureau of Economic Research, Cambridge, MA. Lumsdaine, R. and Papell, D. (1997), ‘Multiple Trend Breaks and the Unit Root Hypothesis’, Review of Economics and Statistics, vol.51, 106–21. M2 Presswire (2004), ‘Report on strategies to improve Africa’s agriculture, food security’, presented at UN Headquarters, Coventry, 28 June [http//:proquest.umi.com]. Mackinnon, J. (1991), ‘Critical Values for Cointegration Tests’, in R. Engle and C. W. J. Granger (eds.), Long Run Economic Relationships: Readings in Cointegration, Oxford: Oxford University Press. Maddala, G. S. (1992), ‘Introduction to Econometrics’, 2nd edn., New York: McMillan. Maizels, A. (1992), Commodities in Crisis: The Commodity Crisis of the 1980s and the Political Economy of International Commodity Policies, Oxford: Clarendon Press. —— (1994), ‘Commodities in Crisis’, chapter 2 in D. Sapsford and W. Morgan (eds.), The Economics of Primary Commodities: Models, Analysis and Policy, Cheltenham: Edward Elgar, 9–29. —— (ed.) (1987). ‘Primary Commodities in the World Economy: Problems and Policies’, World Development, May. Mallick, S. K. (1999), ‘Modelling Macroeconomic Adjustment With Growth in Developing Economies: The Case of India’, Ashgate. Matthews, A., Bureau, J. C., and Jean, S. (2005), ‘Agricultural trade liberalization: assessing the consequences for developing countries’, Contributed paper to the 11th Congress of the European Association of Agricultural Economists, Copenhagen, 24–7 August. Mayer, G. and Wood, A. (2001), ‘Africa’s Export Structure in a Comparative Perspective’, Cambridge Journal of Economics, 25 (3), 369–94. Mbelle, A. (2002), ‘HIPC Relief: Too Late, Too Little?: Perspective from a New Qualiﬁer’, WIDER Discussion Paper No. 2002/122, United Nations University. McCorriston, S., Sexton, R. J., and Sheldon, I. M. (2005), ‘Vertical Market Structure, Commodity Exports and Trade Reform’, Paper presented at the XIth conference of the European Agricultural Economics Association, Copenhagen, Denmark 2005. McDonald, D. (1982), ‘Debt Capacity and Developing Country Borrowing: A Survey of the Literature’, IMF Staff Papers, December 1982. McMillan, M. S. and Masters, W. A. (2000), Understanding the political economy of agriculture in the tropics’, American Journal of Agricultural Economics 82, 738–42. Meier, G. M. (1968), ‘The International Economics of Development: Theory and Policy’, Harper and Row, New York.

337

References Milner, C., Morrissey, O., and Rudaheranwa (2000), ‘Policy and Non-Policy Barriers to Trade and Implicit Taxation of Exports in Uganda’, Journal of Development Studies, 37 (2), 67–90. Morisset J. P. (1998), ‘Unfair Trade? The Increasing Gap between World and Domestic Retail Prices in Commodity Markets during the past 25 years’, The World Bank Economic Review 12 (3) 503–26. Morrissey, O. (2002), ‘Making Debt Relief Conditionality Pro-Poor’, WIDER Discussion Paper No. 2002/04, United Nations University. Mundlak, Y. and Larson, D. F. (1992), ‘On the Transmission of World Agriculture Prices’, The World Bank Economic Review, 6 (3), 399–422. Muscatelli, V. A. (1995), ‘NIE Export Performance Revisited: The Estimation of Export Demand Elasticites, and the Role of Product Differentiation and Growth’ chapter 2 in Currie D. and Vines, D. (eds.) North–South Linkages and International Macroeconomic Policy, Cambridge University Press, 9–28. Nelson, C. R and Plosser, C. I. (1982), ‘Trends and Random Walk in Macroeconomic Time Series: Some Evidence and Implications’, Journal of Monetary Economics, 10, 139–62. Newbery, D. and Stiglitz, J, (1981), ‘Summary of Findings’, chapter 3 in The Theory of Commodity Price Stabilization: A Study in the Economics of Risk, Oxford: Oxford University Press, 23–46. Newbold, P. and Vougas, D. (1996), ‘Drift in the Relative Price of Primary Commodities: a Case Where We Care About Unit Roots’, Applied Economics, 28, 653–61. Nicita, A. (2005), ‘Multilateral Trade Liberalization and Mexican Households: The Effect of the Doha Development Agenda’, ch. 4 in T. W. Hertel and L. A. Winters (eds.), Poverty Impacts of a WTO Agreement: Putting Development back into the Doha Agenda [www.worldbank.org/trade/wto]. Nkamleu, G.B. (2004), ‘Productivity Growth, Technical Progress and Efﬁciency Change in African Agriculture’, African Development Review 16 (1), 203–22. Nyangito, H. O. and Karugia, J. T. (2002), ‘The Impact of Recent Policy Changes on the Agricultural Sector and Public Agricultural Research in Kenya’, ch. 5 in D. Bigman (ed.), Globalization and the Developing Countries: Emerging Strategies for Rural Development and Poverty Alleviation, International Service for National Agricultural Research, The Hague. Okunmadewa, F. and Olayemi, J. K. (2002), ‘Globalization, internal policy reforms, and public agricultural research in Nigeria’, ch. 13 in D. Bigman (ed.), Globalization and the Developing Countries: Emerging Strategies for Rural Development and Poverty Alleviation, International Service for National Agricultural Research, The Hague. Oxfam (2001), ‘Debt Relief: Still Failing the Poor’, Oxfam International. Page, S. and Hewitt, A. (2001), ‘World Commodity Prices: Still a Problem for Developing Countries?’, Overseas Development Institute, London. Perkins, D. H., Radelet, S., Snodgrass, D. R., Gillis, M., and Roemer, M. (2001), Economics of Development, 5th edn., New York: Norton. Perron, P. (1988), ‘Testing for a Unit Root in a Time Series With Changing Mean’, Manuscript, Princeton University. —— (1989), ‘The Great Crash, the Oil Price Shock, and the Unit Root Hypothesis’, Econometrica, 57, 1361–1404.

338

References Pesaran, M. H., Shin, Y., and Smith, R. J. (2001), ‘Bounds Testing Approaches to the Analysis of Long Run Relationships’, Journal of Applied Econometrics, Special Issues in Honour of J. D. Sargan on the Theme, Studies in Empirical Macroeconometrics, ed. D. F. Hendry and M. H. Pesaran, 289–326 and also University of Cambridge DAE Working Paper No.9907 [accessed from www.econ.cam.ac.uk/faculty/pesaran/ pss1r1.pdf on 5/2/2001]. Philips, P. C. B. and Hansen, B.E. (1990), ‘Statistical Inference in Instrumental Variables Regression with I (1) Processes’, Review of Economic Studies, 57(1),99–125. Pingali, P. (2005), ‘Trade liberalization and the least developed countries’, Invited paper to the 11th Congress of the European Association of Agricultural Economists, Copenhagen, 24–7 August. Ponte, S. (2001), ‘The ‘Latte’ Revolution’? Winners and Losers in the Restructuring of the Global Coffee Marketing Chain’, Centre for Development Research Working Paper, Copenhagen. Powell, A. (1991), ‘Commodity and Developing Countries Terms of Trade: What Does the Long Run Show?’, Economic Journal, 101, 1485–96. Prebisch, R. (1950). ‘The Economic Development of Latin America and its Principal Problems’, New York: UN Commission for Latin America. —— (1962), ‘The Economic Development of Latin America and its Principal Problems’, Economic Bulletin for Latin America, 7, 1–22. Radetzki, M. (1990), A Guide to Primary Commodities in the World Economy, Oxford: Basil Blackwell. Ramsey, J. K, (1969), ‘Test for Speciﬁcation Errors in Classical Linear Least square regression analysis’, Journal of the Royal Statistical Society, 350–71. Ranis, G. and Stewart, F. (2001), ‘HIPC: Good News for the Poor’, paper presented at the UNU/WIDER Conference on Debt Relief and HIPC initiative. Rao, D. S. P. (1993), ‘Inter-country Comparisons of Agricultural Output and Productivity’, FAO, Rome. Razzaque, M., Grynberg, R., and Osafa-Kwaako, P. (2004), ‘Commodity Prices, Aid and Debt: Implications for LDCs, SVS, and HIPCs’, paper prepared for the Commonwealth Secretariat and UNCTAD. Redding, S. and Venables, A. (2001), ‘Economic Geography and International Inequality’, mimeo: London School of Economics. Reinhart, C. M. and Wickham, P. (1994), ‘Commodity Prices: Cyclical Weakness or Secular Decline?’, IMF Staff Papers, 41 (2), 175–213. Roberts, J. and Fagerna¨s, S. (2004), ‘Why is Bangladesh Outperforming Kenya: A Comparative Study of Growth and its Causes Since the 1960s’, ESAU Working Paper No. 5, Economics and Statistics Analysis Unit, Overseas Development Institute, London. Rodrik, D. (1995), ‘Getting Interventions Right: How South Korea and Taiwan Grew Rich?’, Economic Policy, 55–107. Rogers, C. R. (1973), ‘International Commodity Agreements’, Lloyds Bank Review, April. Roodman, D. (2001), ‘Still Waiting for the Jubilee: Pragmatic Solutions for the Third World Debt Crisis’, Washington, D.C.: World Watch Institute. Sachs, J. et al. (2004), ‘Ending Africa’s Poverty Trap’, Papers on Economic Activity 1, Brookings Institution, Washington, D.C.

339

References Sapsford, D. R. (1985), ‘The Statistical Debate on the Net Barter Terms of Trade Between Primary Commodities and Manufactures: A Comment’, Economic Journal, 95, 781–8. Sapsford, D.R. and Balasubramanyam, V. N. (1994), ‘The Long-run Behaviour of the Relative Price of Primary Commodities: Statistical Evidence and Policy Implications’, World Development, 22, 16–70. —— and Morgan, C. W. (eds.) (1994), The Economics of Primary Commodities: Models, Analysis and Policies, Edward Elgar: Aldershot, UK. —— and Singer, H. (1998), ‘The IMF, the World Bank and Commodity Prices: A Case of Shifting Sands?’, World Development, 26 (9), 1653–60. —— Sarkar, P., and Singer, H. (1992), ‘The Prebisch–Singer Terms Of Trade Controversy Revisited’, Journal of International Development, 4 (3), 315–32. Sarkar, P. (1986), ‘The Singer-Prebisch Hypothesis: A Statistical Evaluation’, Cambridge Journal of Economics, December. —— (1986). ‘The Terms of Trade Experience of Britain in the Nineteenth Century’, Journal of Development Studies, 23, 20–39. —— and Singer, H. (1991), ‘Manufactured Exports of Developing Countries and Their Terms of Trade Since 1965’, World Development. Sawada, Y. (2003), Immiserizing Growth: An Empirical Evaluation, Unpublished paper, Graduate School of Economics, University of Tokyo. Schlote, W. (1938; 1952 edn.), British Overseas Terms of Trade from 1700 to 1930s, Oxford: Basil Blackwell. Senhadji A. and Montenegro, C. (1998), ‘Time Series Analysis of Export Demand Equations: A Cross-Country Analysis’, IMF Working Paper, WP/98/149. Serieux, J. (2001), ‘The Enhanced HIPC Initiative and Poor Countries: Prospects for a Permanent Exit’, Canadian Journal of Development Studies, 22, (2), 527–48. Shepard, B. (2004), ‘Market Power in International Commodity Processing Chains: Preliminary Results from the Coffee Market’, Groupe D’Economie Mondiale, Sciences Po. Shikwati, J. S. and Okonski, K. (2005), ‘Will the Destruction of Coffee Stocks save Kenyan Farmers?’ Inter Region Economic Network, Nairobi. [http://irenkenya.org/ page.php?instructions¼page&page_id¼45&nav_id¼19]. Singer, H. (1950), ‘The Distribution of Gains Between Investing and Borrowing Countries’, American Economic Review: Papers and Proceedings, 40, 473–85. —— and Edstrom, J. (1993), ‘The Impact of Trends and Volatility in Terms of Trade on GNP Growth’, Chapter 1 in Nissanke, M., and Hewitt, A. (eds.), Economic Crisis in Developing Countries: New Perspectives on Commodities, Trade and Finance, London: Printer Publishers, 23–45. Spraos, J. (1980), ‘The Statistical Debate on the Net Barter Terms of Trade Between Primary Commodities and Manufactures’, Economic Journal, 90, 107–28. Stevens, C., McQueen, M., and Kennan, J. (1999), ‘After Lome´ IV: A Strategy for ACP-EU Relations in 21st Century’, Economic Paper 37, Commonwealth Secretariat, London. Stiglitz, J. E. (2000), ‘Two Principles for the Next Round: How to Bring Developing Countries From the Cold’, The World Economy, 437–54. Talbot, J. (1997), ‘Where Does Your Coffee Dollar Go?: The Division of Income and Surplus along the Coffee Commodity Chain’, Studies in Comparative International Development, 32(1), 56–91.

340

References —— (2002), ‘Tropical Commodity Chains, Forward Integration Strategies and International Inequality: Coffee, Cocoa and Tea’, Review of International Political Economy, 9(4), 701–34. Thirlwall, A. P. (1989), ‘Growth and Development With Special Reference to Developing Countries: Fourth Edition’, ELBS, MacMillan, London. —— (1999). Growth and Development: With Special Reference to Developing Countries, 6th edn., London: Macmillan. Thirlwall, A. P. and Bergevin, J. (1985), ‘Trends, Cycles, and Asymmetries in the Terms of Trade of Primary Commodities from Developed and Less Developed Countries’, World Development, July. Timmer, M. and Los, B. (2005), ‘Localized Innovation and Productivity Growth in Asia: An Intertemporal DEA Approach’, Journal of Productivity Analysis, 23(1), 47–64. UNCTAD (1996), ‘The Least Developed Countries 1999 Report’, UNCTAD, Geneva. —— (1997), ‘The Least Developed Countries 1999 Report’, UNCTAD, Geneva. —— (1998), ‘The Least Developed Countries 1999 Report’, UNCTAD, Geneva. —— (1998), ‘Trade and Development Report 1998’, United Nations, Geneva. —— (1999), ‘The Least Developed Countries 1999 Report’, UNCTAD, Geneva. —— (2000), ‘Debt Relief, the New Policy Conditionality and Poverty Reduction Strategy’ Chapter 5 in The Least Developed Countries 2000 Report, 135–70, Geneva. —— (2001), World Investment Report 2001: Promoting Linkages, United Nations, New York and Geneva. —— (2002a), ‘The Least Developed Countries Report 2002: Escaping the Poverty Trap’, United Nations, Geneva. —— (2002b), ‘UNCTAD Handbook of Statistics 2002’, United Nations, Geneva. —— (2002c), ‘Export Diversiﬁcation, Market Access and Competitiveness’, Report by the UNCTAD Secretariat, TD/B/COM.1/54. —— (2004a), Haque, ‘Commodities under Neo-Liberalism: The Case of Cocoa’, G-24 Discussion Paper Series. —— (2004b), ‘Economic Development in Africa: Trade Performance and Commodity Dependence’, United Nations Publication, sales no. E.03.II.D.34. New York and Geneva. —— and Commonwealth Secretariat Joint Study (2001), ‘Duty and quota-free market access for LDCs: an analysis of quad initiatives’, UNCTAD/DITC/TAB/Misc.7, London, England and Geneva, Switzerland. United Nations (1949), Relative Prices of Exports and Imports of Under-Developed Countries, New York. Von Braun, J., Boue¨t, A., Cororaton, C. B., Mengistu, T., and Orden, D. (2005), ‘Trade policies to address vulnerability of the poor in developing countries and opportunities to complementary action’, Invited paper to the 11th Congress of the European Association of Agricultural Economists, Copenhagen, 24–7 August. Vorley, B. (2003), ‘Food, Inc., Corporate Concentration from farm to consumer’, UK Food Group. White, H. (1980), ‘A Heteroscedasticity Consistent Variance Matrix Estimator and a Direct Test of Heteroscedasticity’, Econometrica, 48, 817–18. —— (2002), ‘Long-run Trends and Recent Developments in Ofﬁcial Assistance from Donor Countries’, WIDER Discussion Paper No. 2002/106.

341

References Wilson, T., Sinha, R. P., and Castree, J. R. (1969), ‘The Income Terms of Trade of Developed and Developing Countries’ Economic Journal, December. Winters, L. A. (2005), ‘The European Agricultural Trade Policies and Poverty’, European Review of Agricultural Economics, 32(3), 319–46. —— and Martin, M. (2003): ‘Small is Beautiful but Costly: Business Costs in Small Economies’ mimeo: Commonwealth Secretariat, London. —— Walmsley, T. L., and Wang, Z. K. (2002), ‘Negotiating the Liberalisation of the Temporary Movement of Natural Persons’, Commonwealth Secretariat. —— McCulloch, N. and McKay, A. (2004), ‘Trade liberalization and poverty: the evidence so far’, Journal of Economic Literature, 42 (1), 72–115. —— and Sapsford, D. (1990), ‘Primary Commodity Prices: An Introduction to the Major Policy and Modelling Challenges’, in L. A. Winters and D. Sapsford (eds.), Primary Commodity Prices: Economic Models and Policy, Cambridge: Cambridge University Press. —— Walmsley, T., Wang, Z. K., and Grynberg, R. (2002), ‘Liberalising Labour Mobility Under the GATS’, Economic Paper No. 53, Commonwealth Secretariat, London. Wood, A. and Jordan, K. (2000), ‘Why Does Zimbabwe Export Manufactures and Uganda Not: Econometrics Meets History’, Journal of Development Studies, 37 (2), 67–90. —— and Mayer, J. (2001), ‘African Export Structure in a Comparative Perspective’, Cambridge Journal of Economics, 25, 3, 369–94. World Bank (1993), ‘Global Economic Prospects and the Developing Countries 1993’, World Bank: Washington, D.C. —— (1994), ‘Global Economic Prospects and the Developing Countries 1994’, World Bank: Washington, D.C. —— (1996), ‘Global Economic Prospects and the Developing Countries 1996’, World Bank: Washington, D.C. —— (1997), ‘Global Economic Prospects and the Developing Countries 1997’, World Bank: Washington, D.C. —— (1998), Assessing Aid: What Works, What Doesn’t and Why, Oxford: Oxford University Press. —— (1999). ‘Dealing With Commodity Price Volatility in Developing Countries: A Proposal for Market Based Approach’, Discussion Paper for the Roundtable on Commodity Risk Management in Developing Countries, World Bank: Washington, D.C. —— (2000), World Development Indicators, CD-Rom, World Bank: Washington, D.C. —— (2002), Global development ﬁnance, Economic Policy and Prospects Group, Washington, D.C. [www.worldbank.org/prospects/gdf2002/]. —— (2003), ‘Global Development Finance’, World Bank: Washington, D.C. —— (2003), ‘Kenya: Exporting out of Africa—Kenya’s Horticultural Success Story’, Economic Policy and Prospects Group, Washington, D.C. [www.worldbank.org/ wbi/reducingpoverty/case-Kenya-Horticultural.html]. —— (2004), ‘Agricultural Diversiﬁcation for the Poor, Guidelines for Practitioners’, World Bank ARD Discussion Paper 1, 2004. —— and IMF (2003), ‘Enhanced HIPC Initiative—Creditors Participation Issues’, 8 April, accessed from www.worldbank.org on 22 August 2003. WTO (2003), ‘Non Paper on the Need for Urgent Action in WTO to Deal With the Crisis Situation Created by the Long-term Trend Towards Decline in Prices of Primary Commodities to the Trade and Development of Developing Countries Which are

342

References Heavily Dependent on Their Exports’, Committee on Trade and Development, WT/ COMTD/W/113, dated 19 May. Yabuki, N. and Akiyama, T. (1996), ‘Is Commodity-Dependence Pessimism Justiﬁed? Critical Factors and Government Policies that Characterize Dynamic Commodity Sectors’, Policy Research Working Paper 1600, World Bank: Washington, D.C. Zivott, E., and Andrews, D. (1992), ‘Further Evidence on the Great Crash, the Oil price Shock, and the Unit Root Hypothesis’, Journal of Business and Economic Statistics, 10, 251–70.

343

This page intentionally left blank

Index

adding-up problem 11 Addison, T. 312 Afghanistan 189–96 African Development Bank 286 Agenor, P. R. 301 agriculture: causes of stagnation 69–70 global export share 10–11 output index 93–4 productivity change 81–7 surge in production 28 aid 4, 163, 299 and commodity prices 302–6 fungibility 310–11 ODA 227, 230 Akiyama, T. 173 Alauddin, M. 86, 93, 105 alumina 275 Amin, A. A. 108 Amjadi, A. 226 Andersen, R. W. 272 Andrews, D. 23, 25, 27 Angola 198, 286 Antigua 205, 206 Appleyard, D. R. 87 Ardeni, P. G. 25 Argentina 105 Athukorala, P. C. 18, 42, 219 Atkins, J. P. 176 Australia 144 Australia-US, farm gate-to-retail spread 148, 149, 150, 158 Baanante, C. 109 Bahrain 178, 206 Balasubramanyam, V. N. 17 Balat, J. F. 109 bananas 39–67, 276 Bangladesh 104–5, 111, 115, 117–18, 189–96, 198, 202, 208 Barbados 79, 107–8, 115 Barbuda 205–6 bauxite 275 beef/veal 276

Belgium 198 Benin 104, 173, 189–96, 288 Bera, A. K. 222, 225 Bergevin, J. 20–1 Berlage, L. 312 Bernall, R. 226 Beveridge, S. 23, 25 Bhagwati, J. 77–8 Bhutan 189–96 Birdsall, N. 282, 287, 291, 295–6, 298 Bleaney, M. F. 28, 35–8, 41, 43, 49–50, 279 Bloch, H. 29 Bolivia 81, 171, 288 Borensztein, E. 28 bottlenecks 141 Boue¨t, A. 100 Braga, C. P. 183 Brazil 105, 144 Brazil-US, farm gate-to-retail spread 148, 149, 150, 157 buffer stocks 269, 270 Burkina Faso 104, 109, 288, 292 Burundi 72, 171, 189–96 Cambodia 202, 205 Cameroon 101, 107, 108, 122, 143, 292 Cameroon-UK, farm gate-to-retail spread 146, 151 Canada 196, 198 Cape Verde 207 Cashin, P. 26 Cassen, R. 302, 310 catch up 86 Central African Republic 123–4, 171, 173 Chad 104, 123, 124, 173, 292 Chang, C. 302 Chile 89, 103, 105 China 196, 198 Choraria, J. 79, 101 civil unrest 107–8, 129, 189, 196, 230 cobalt 275 cocoa 39–67, 72, 130–4, 274, 282, 302 and aid 302, 305

345

Index cocoa (cont.) farm gate-to-retail spread 147, 153 foreign exchange loss 173 prices: fob 140; producer 101, 144; trends 39–67 coconut oil 173 Coelli, T. 82–3, 93 coffee 39–67, 72, 79, 101, 107, 130, 274, 282 farm gate-to-retail spread 145–6, 147, 151–3 prices: fob 140; producer 101, 143; trends 9–67; UK retail 144 value chain 138–9 cointegration techniques 217–8, 222–3 Colombia 123, 125 commercial services: by country 239–40, 243–44, 249–52, 261–3 exports 183–6, 198, 205 commodities: declining use of 28, 208 dependence on 7–10 mix 109–10 commodity prices: and aid ﬂow 302–6 classical view of 18 and debt 279–82 and Heavily Indebted Poor Countries (HIPC) Initiative 287–303 see also prices; relative prices commodity value chain: data 142–5 deﬁnition 137 the literature 137–42 and market power 141–2 Common Fund for Commodities (CFC) 272 Commonwealth countries 69, 72, 74–7, 105–9 Comoros 81 Congo DR 198 consumption shocks 139 convergence variable 112 copper 39–67, 275 copra 274 Cororaton, C. B. 100 Costa Rica 126 costs: changes in 139 transportation 226–7 ˆ te d’Ivoire 72, 104, 171, 173, 274, 282 Co Cotonou agreement 115, 274–5 cotton 39–67, 72, 85, 274, 282, 302–13 foreign exchange loss 173 productivity 109–12 relative price trends 39–67 Cuddington, J. T. 22–5, 27–8, 50 Cypher, J. M. 78–9 Cyprus 184

346

Daseking, C. 282 Deaton, A. 29, 39, 165, 279 debt 3 and commodity prices 279–82 net present value (NPV) 285, 286, 287, 290 service payments 292–4 sustainability 282, 285, 287, 291–2 top-up relief 293, 295–8 see also Heavily Indebted Poor Countries Deese, B. 295–6, 298 DEFRA 140–1 Dehn, J. 279 demand 69 income elasticity of 11, 18, 28, 78, 210 Denmark 313 Devarajan, S. 310 Diakosavvas 19–21 Dickey, D. A. 39, 41 Dietz, J. L. 78–9 Djibouti 206 Dodhia, D. 285 Dominica 173 Douya, E. 108 Duncan, R. C. 114 Dutch disease 11, 106, 108 Easterly, W. 279–80 Eastern Europe 28 Economist’s index of industrial commodity prices 26 economy: importance of selected commodities 72–4 inefﬁciencies 229–30 edible oils 173 Edstrom, J. 279 Engel’s Law 208, 210 Engle, R. F. 23, 26, 37, 140, 215 Equatorial Guinea 178, 198, 205, 207 error correction model 36, 140 Ethiopia 143 Ethiopia-UK, farm gate-to-retail spread 146 European Commission 69, 72, 78 European Union (EU) 229 ACP Programmes 3, 115, 147–8, 275–6; COMPEX 272, 275; STABEX 3, 272–4, 306–7, 313; SYSMIN 272, 275 Common Agricultural Policy (CAP) 142, 276 Common Market Organization (CMO) 147 Special Preferential Sugar (SPS) Agreements 148 export growth 71 by country 75–7, 194–5, 235–6 long-term 189–96 projected 288–90 trend coefﬁcients 73–4

Index export unit value 71 across country groups 94–100 index 94 and TFP 113–22 and world import unit values 98–100 exports: by country 89–91, 190–1, 233–4 commercial services 183–6, 198, 205 prices 70, 142, 144–5 quantities 74–7 quotas 270 of services 163, 183–6, 194–5, 198, 205, 233–4 shares 192–3 structural shift 299 total 186–7, 198–200, 253–6, 264–5 value 171 volume 69, 167, 177 see also merchandize exports External Compensatory Finance 269, 272–5, 306

Gillson, I. 293 globalization 211, 229 Godfrey, L. G. 222, 225 gold 275 governance structures 137 government role 74 Granger, C. W. J. 23, 26, 37, 140, 215 Greece 198 Greenaway, D. 17, 28, 35–8, 41, 43, 50, 279 Grenadines 207 Grilli, E. R. 19, 21–6, 28, 36, 39–40, 43, 279 Grilli-Yang dataset 40–4, 50–1, 57, 63–5 gross domestic product (GDP), per capita 85, 111–12 groundnuts 274 Grynberg, R. 136, 229, 292 Guatemala 123, 125 Guinea Bissau 171, 288 Gujarati, D. N. 26 Guyana 81

Fagerna¨s, S. 105, 119 Field, A. J. 87 ﬁeld crops 72, 74, 75 Fiji 79, 107–8, 115, 144, 173, 207 Fiji-EU, farm gate-to-retail spread 149, 156 Fiji-US, farm gate-to-retail spread 148–50, 156 ﬁnance, external compensatory 269, 272–5, 306 ﬁnancial market derivatives 276 Fitter, R. 139, 141, 146 foreign direct investment 228 foreign exchange loss 2–3 by commodities 172–3 by country 174 methodology 163–7 results 167–71 France 196 Fuglie, K. O. 103 Fuller, W. A. 39, 41

Hadass, Y. 26 Haiti 189–96, 208 Hall, S. 216 Hansen 218 Harris, R. 26, 216 Harvey, A. C. 25 Havana Charter (1948) 270 Headey, D. 86 Healy, S. 139 Heavily Indebted Poor Countries (HIPC) Initiative 3–4, 278, 282–6 commodity export dependence 7–10 and commodity prices 287–93 conditionality 285, 292 costs 286, 296–8 countries listed 13–14 criticised 287 leading exports 15–16 real price adjustment mechanism 295–8 Henao, J. 109 Hendry, D. F. 22 Herrmann, R. 269, 274 Hertel, T. W. 100 Hewitt, A. 274–6, 293, 306–7, 313 high-performing Asian economies (HPAE) 181 Hoekman, B. 183 Holden, D. 219 Honduras 81, 173, 288, 293 Hong Kong 196, 198 Hopkins, T. K. 137

Gabon 104, 115, 117, 126, 171, 178, 206 Gallup, J. L. 85–6, 89 Gambia 81, 107, 108, 189–96 Gautam, M. 280, 282, 285 Geda, A. 280 General Agreements of the Trade in Services (GATS) 183 Gerefﬁ, G. 137 Germany 196 Ghana 72, 79, 101, 107–8, 118–19, 173 Ghana-UK, farm gate-to-retail spread 147, 153 Gibbon, P. 137 Gilbert, C. L. 272 Gilbert, R. 295 Gillis, M. 88

IAC 69 IADB 107 IDA 290 illiteracy rate 86, 111

347

Index immiserizing growth 71 evidence of 79–80 mechanics of 77–9 see also poverty import-substitution strategy 226, 293 income distribution, commodity value chain 137 income elasticity of demand 11, 18 low 28 manufactures 78, 210 indebtedness see debt India 119, 198 Indonesia 72, 103, 123, 125, 196 infrastructure, social and physical 227–8 institutional framework 137 insurance 226, 277 Integrated Programme for Commodities (IPC) 272 International Cocoa Organization (ICCO) 144 International Coffee Agreement (ICoA) 272 International Coffee Organization (ICO) 138, 143–4 International Commodity Agreements (ICAs) 69, 139, 269–72 International Monetary Fund (IMF) 3, 30, 79, 114, 119, 212, 282, 286, 290 Compensatory Financing Facility (CFF) 272–3, 306 International Natural Rubber Agreement (INRA) 272 International Sugar Agreement (ISA) 270 International Tin Agreement (IAT) 40 Ireland 196, 198 iron ore 275 Jamaica 107, 123, 124, 184 Japan 196, 313 Jarque, C. M 222, 225 Joint Diversiﬁcation Fund (JDF) 306–14 costs of 311–4 country allocations 316–27 design of 307–11 jute 39–67 Kanbur, R. 309 Kaplinsky, R. 139, 141, 146 Karugia, J. T. 119 Kellard, N. 27 Kenya 81, 101, 104, 105, 119, 129, 143, 173, 286 Kenya-UK, farm gate-to-retail spread 146, 152 Keynes, J.M. 269 Kiribati 189–96 Kiringai, J. 292 Knapp, R. 134 Korzeniewicz, M. 137 Kuhnen, F. 78

348

labour absorption 105 labour productivity 85–6 changes in; causes 111–13; rate of 71, 101–2 deﬁnition 81 and TFP 104–5 see also productivity Lao PDR 81, 189–96, 202 Laroque, G. 29 Larson, D. F. 140 lauric oils 72 Leon, J. 24–5 Lesotho 205 less developed countries (LDCs), list of 13–14, 231 Lewis, W. A. 21, 29 Limao, N. 227 Lipsey, R. E. 114 literature review 17–34 econometric models 21–8 structural models 28–9 to mid-1980s 20–1 LMC International 144 Lome´ Conventions 115, 147, 229, 274–5, 313 Lumsdaine, R. 27 Luxembourg 313 McCorriston, S. 79 McCulloch, N. 100 McDermott, C. J. 26 McKay, A. 100 Madagascar 124, 125, 208 maize 39–67 Maizels, A. 49, 164–5, 270 Malawi 111, 126 Malaysia 104, 107, 119–20, 196 Maldives 205 Mali 104, 126, 173, 293, 302–3 Mallick, S. K. 219 Malta 184, 198 manganese 275 manufacturing base 226 marginalization 176, 196 avoidance 199–205 by country 197 data 212–3 and development 225–30 estimation of 205–9 LDCs and SVSs 181–3 long- and short-run relationships in 219–21, 224 and merchandise export trade 208–25 model estimation 213–24 simple model 211 market power 141–2 market structure 18 Mauritania 104, 290 Mauritius 101, 107, 126, 130, 144, 184

Index Mauritius-EU, farm gate-to-retail spread 149, 155 Mauritius-US, farm gate-to-retail spread 148, 149, 150, 155 Mayer, J. 227 Mbeaoh, A. 108 Mengistu, T. 100 merchandise exports: by country 237–8, 241–2, 245–8, 257–60 change in 196 LDCs and SVSs 176, 178, 180–3 marginalization in 208–25 see also exports merchandise imports 178–9 Mexico 196 Miller, R. 165 Montenegro, C. 219 Montiel, P. J. 301 moral hazard 310–11 Morgan, C. W. 17 Morisset, J. P. 136, 138–42, 146 Morrissey, O. 285 Mozambique 107–8, 122, 126, 288 Mundlak, Y. 140 Muscatelli, V. A. 219 Myanmar 81, 171, 189–96, 198, 202 National Statistics Ofﬁces 144 natural rubber 173 Nelson, C. R. 22–3, 25 Nepal 104–5, 189–96, 198, 202, 206 Netherlands 196, 313 Newbery, D. 306 Newbold, P. 23, 27 Nicaragua 81 Nicita, A. 100 Niger 104, 171, 189–96, 292 Nigeria 72, 106, 119, 124 Nkamleu, G. B. 82–3, 86 Norway 313 Nyangito, H. O. 119 oil-exporters 178, 180, 181, 188 Okonski, K. 129 Okunmadewa, F. 106 Olayemi, J. K. 106 openness index 111 Orden, D. 100 Osafa-Kwaako, P. 136 Page, S. 275–6, 306–7, 313 palm oil 39–67, 72, 173, 282 Papell, D. 27 Papua New Guinea 79, 107, 111, 124–5, 143, 173, 206, 302, 305 Papua New Guinea-UK, farm gate-to-retail spread 146, 152

Paris Club 286 Perkins, D. H. 88 Perman, R. 219 Perron, P. 23, 25, 27 Pesaran, M. H. 36–8, 219–20, 223 Philippines 196 Phillips, P. C. B. 218 phosphates 275, 302, 305–6 physical conditions 86, 112–13 Plosser, C. I. 22 policy change 139 policy instruments 3, 269–77 Ponte, S. 137, 139, 141 population, rural proportion 86, 111 Porto, G. G. 109 poverty: alleviation 292, 293 extreme 12–13 see also immiserizing growth Poverty Reduction Growth Facility (PRGF) 273 Powell, A. 25, 27, 37, 50 Powell, R. 282 Prebisch, R. 1, 7, 17–20, 22, 28, 35, 78–9 preferential trade agreements 270, 275–6 price elasticity 11, 18 price index 94, 100–1, 109 price shocks 11, 74, 189 prices 109, 210 adjustment mechanism 295–8 asymmetric transmission 140–1 commodity-retail spread 138–9, 145–7, 149–50 exports 70; fob 142, 144–5 producer 70, 94, 100–1, 142, 144–7 and productivity real 82 see also commodity prices; relative prices producer price index production data 91–4 productivity 210 cross country and sectoral 70–1 and prices 71, 78, 85, 87–8 role in production 68–9 see also labour productivity; total factor productivity purchasing power index 167, 171 purchasing power parity (PPP) 11 quality improvements 19 Radelet, S. 88 Radetzki, M. 307 Ramsey, J. K. 222, 225 Ranis, G. 293 Rao, D. S. P. 82–3, 86, 93 Razzaque, M. A. 136, 292 Redding, S. 227

349

Index Reinhart, C. M. 25, 28, 210 relative prices 1–2, 11, 17–34 for commodity groups 43–7 growth rate 40–3 individual commodities 47–51 methodology 36–9 model 37–9 trends 39–67 see also commodity prices; prices remittances from abroad 163 Republic of Korea 196 rice 72, 85, 110–11, 39–67 Riedel, J. 219 risk management instruments 270, 276–7 Roberts, J. 105, 119 Roemer, M. 88 rubber 282 rum 276 Russian Federation 196 Rwanda 72, 189–96, 293 Sachs, J. D. 85–6, 89 St Kitts and Nevis 206 St Lucia 173, 206 St Vincent 206 Samoa 173 Sao Tome and Principe 189–96, 206, 282 Sapsford, D. 17, 20, 24, 27–9, 50, 173 Sarkar, P. 18–21 Saudi Arabia 196 Sawada, Y. 78–9 Scandizzo, P. L. 19–21 Schlote, W. 21, 24 Senegal 274, 288, 293 Senhadji, A. 219 Serieux, J. 292 Sexton, R. J. 79 Sheldon, I. M. 79 Shepard, B. 139–40 Shikwati, J. S. 133 shocks: consumption 139 external 275, 291, 295 price 11, 74, 189 supply 189 Sierra Leone 107–8, 122, 129, 189–96, 208, 292 Singapore 196 Singer, H. 1, 7, 17–20, 22, 28, 35, 78–9, 173, 279 small vulnerable states (SVS), listed 13–14, 232 Snodgrass, D. R. 88 soil quality 112–13 Solomon Islands 107, 115, 117, 126, 127, 173 Soto, R. 24–5 South Africa 89, 106 South Korea 196 Soviet Union 28 Spain 196, 198

350

Spraos, J. 19–20 Sri Lanka 107, 119, 126, 128 stability, social and political 229 Stewart, F. 293 Stiglitz, J. E. 228, 306 structural break 50 sub-Saharan Africa 69 Sudan 173, 198 sugar 72, 85, 110–12, 130–4, 275–6 farm gate-to-retail spread 154–7 policies 147–8 prices: farm gate 144; fob 144–5, 148; retail 145; trends 39–67 Sugar Protocol 2, 147, 148, 150 supply shock 189 Suriname 189–96, 206 Swaroop, V. 310 Swaziland 79 Sweden 313 Tabova, A. 295 Taiwan 196 Talbot, J. 138–40 Tanzania 79, 107, 119, 143, 171, 287–8, 293 Tanzania-UK, farm gate-to-retail spread 146, 153 tea 39–67 technical progress 18, 81, 208, 210 technological change differentials 78 terms of trade 17, 19 circular deterioration 78 net barter (NBTT) 18, 23, 114 single factoral 71, 87–8, 122–34 Thailand 144, 196 Thailand-US, farm gate-to-retail spread 148, 149, 150, 157 Thirwall, A. P. 20–1 tin 39–67, 275 Tisdell, C. 105 Togo 104, 124, 173, 302, 305–6 Tonga 107, 108, 122, 206 total factor productivity (TPF) 2 between Commonwealth countries 105–9 change: causes of 83–7, 109–11; rate of 71, 102–3 deﬁnition 80–1 and export unit values 113–22 and labour productivity 104–5 model 82–3 national level 70 see also productivity tourism 184, 189 trade barriers 141, 226 trade policy, import-substitution 226, 293 trade transactions by country groups 187–8 trading arrangements, preferential 228–9 tree crops 72, 74–5, 85, 109–11, 122

Index Trinidad and Tobago 79, 107, 115, 122, 130, 178, 189–96 Turkey 89 Uganda 72, 107, 122, 171, 173, 205, 282, 288, 291, 293 unit root tests 36, 214–7, 222–3 United Arab Emirates 196 United Kingdom 18–19, 147, 196, 198 United Nations 69, 114 Food and Agriculture Organization (FAO) 88, 93–4 Commodity Yearbook 172, 212 UNCTAD 2, 68, 164–5, 270, 287 Commodity Price Bulletin 47–51 database 40, 43, 45–7, 51–7, 66–7 United States 196, 198, 313 services supplier 184 sugar export 147 Tariff Rate Quota (TRQ) 148 uranium 275 Uruguay Round 229, 183 Urzua, C. M. 22–5, 28, 50 Vanuatu 107, 108, 122 Venables, A. J. 227 Venezuela 115, 117

Vietnam 81, 170–1, 173, 286 von Braun, J. 100 Vorley, B. 139, 141 Vougas, D. 23, 27 Wallerstein, I. 137 wheat 39–67 White, H. 222, 225, 302 Wickham, P. 25, 28, 210 Williamson, J. 26, 295–6, 298 Wilson, T. 20 Winters, L. A. 100 Wohar, M. E. 27 Wood, A. 227 World Bank 30, 79, 119, 138, 212, 277, 282, 286, 290 World Trade Organization (WTO) 229, 295 Wright, B. 25 Yabuki, N. 173 Yang, M. C. 19, 21–6, 28, 36, 39–40, 43, 279 Yeats, A. 226 Yemen 81, 198, 205, 286 Zambia 107, 108–9, 171, 189–96, 288, 292 Zivott, E. 23, 25, 27

351